Relatively Speaking
Oxford Studies in Anthropological Linguistics William Bright, General Editor
Editorial Board Wai...
29 downloads
673 Views
8MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Relatively Speaking
Oxford Studies in Anthropological Linguistics William Bright, General Editor
Editorial Board Wai lace Chafe, University of California, Santa Barbara; Regna Darnell, University of Western Ontario; Paul Friedrich, University of Chicago; Dell Hymes, University of Virginia; Jane Hill, University of Arizona; Stephen C. Lcvinson, Max Planck Institute, The Netherlands; Joel Sherzer, University of Texas, Austin; David J. Parkin, University of London; Andrew Pawley, Australian National University; Jet' Verschucren, University of Antwerp
Recent Volumes Published 7 Charles L. Briggs (ed.): Disorderly Discourse: Narrative, Conflict, and Inequality 8 Anna Wierzbicka: Understanding Cultures through Their Kev Words: English. Russian. Polish, German, and Japanese 9 Gcrrit J. van Enk and Lourcns de Vries: The Korowai of Irian Jayct: Their Language in Its Cultural Context 10 Peter Bakker: A Language of Our Own: The Genesis of Miehif, the Mixed Cree-French Language of the Canadian Metis 11 Guntcr Senft: Referring to Space: Studies in Austronesian and Papuan Languages 12 David McKnight: People, Countries, and the Rainbow Serpent: Systems of Classification among the Lardil of Mornington Island 13 Andree Tabouret-Keller, Robert B. Le Page, Penelope Gardner-Chloros, and Gabrielle Varro (eels.): Vernacular Literacy Revisited 14 Steven Roger Fischer: Rongorongo, the Easter Island Script: History, Traditions, Text 15 Richard Feinberg: Oral Traditions of'Anuta: A Polynesian Outlier in the Solomon Islands 16 Bambi Schieffelin, Kathryn A. Woolarcl, and Paul Kroskrity (eds.): Language Ideologies: Practice and Theory 17 Susan LI. Philips: Ideology in the Language of Judges: flow Judges Practice Law, Politics, and Courtroom Control 1X Spike Gildca: On Reconstructing Grammar: Comparative Cariban Morphosyntax 19 Laine A. Bcrman: Speaking through the Silence: Narratives, Social Conventions, and Power in Java 20 Cecil H. Brown: Lexical Acculturation in Native American Languages 21 James M. Wilce: Eloquence in Trouble: The Poetics and Politics of Complaint in Rural Bangladesh 22 Peter Seitel: The Powers of Genre: Interpreting Haya Oral Literature 23 Elizabeth Keating: Power Sharing: Language, Rank, Gender, and Social Space in Pohnpei, Micronesia 24 Theodore B. Fernald and Paul Platero (eds.): The Athabaskan Languages: Perspectives on a Native American Language Family 25 Anita Puckett: Seldom Ask, Never Tell: Labor and Discourse in Appalachia 26 Eve Danziger: Relatively Speaking: Language, Thought, and Kinship among the Mopan Max a
Relatively Speaking Language, Thought, and Kinship among the Mopan Maya
EVE DANZIGER
OXJORD U N I V E R S I T Y PRESS
2001
OXJORD UNIVERSITY PRESS Oxford New York Athens A u c k h m d Bangkok Bogota Bombay Buenos Aires Calculia Cape Town Chemuii Dares Salaam Delhi Florence I long Kong I s t a n b u l Karachi Knaki L u m p u r Madrid Melbourne Mexico City Mumbai Nairobi Paris Sao Paulo Shanghai Singapore Taipei Tokyo Toronto Warsaw and associated companies in Berlin Ibadun
Copyright © 2001 by Eve Danzigcr Published by Oxford University Press. Inc. 19S Madison Avenue. New York, New York 10016 Oxford is a registered trademark of Oxford University Press All rights reserved. No pan of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior permission of Oxford U n i v e r s i t y Press. Library of Congress Catuloging-in-Publication Data Dan/iger. Hve. Relatively speaking : language, thought, and k i n s h i p among the Mopan Maya / by I->e Dan/iger. p. cm. — (Oxford studies in anthropological iiiigiiislics) Includes bibliographical references and index. ISBN 0-19-509910-9 1. Mopan dialect—-Belize—Toledo District—Semantics. 2. Mopan dialect—Belize—Toledo District—Grammar—Terminology. 3. Mopan dialect- Belize—Toledo District—Etymology—Names. 4. Mopan Indians—Kinship. 5. Mopan Indians—-Social life and customs. 6. Language and culture—Beli/.e—Toledo District. 7, Toledo District (Beli/e)—Social life and customs. 1. Title. II. Series. PM.W>9.5.M65 D36 2000 497'.4152—dc21 99-OH6793
9 K765432 I Printed in the United States of America on acid-free paper
ACKNOWLEDGMENTS
B etween 1986 and 1989,1 spent fourteen months with the Maya people of the Toledo district of Belize, Central America, thinking about issues of kinship and t:ik 'respect'. My thanks go first of all to them, for permission to live and work among them and for all that they taught me. Particularly gracious were Lino Sho and Demetero Cowoj, alcaldes in 1988 and 1989, who gave their permission for my sojourn in the community and for my work in the village school. Telesforo Paquiul, the principal, also gave his permission and encouragement. With all of the teachers, he welcomed me into the school community in a way that was extremely supportive, encouraging, and helpful. The officers of the Toledo Maya Cultural Council also offered help and support during the fieldwork period. The Peace Coips Language Training Program in Belize allowed me to profit from the services of its Mopan language consultants, Genovevo Peck and Rodolfo Teck. Sessions with them were always an unmitigated delight. The household of Dionisio Bol and Florencia Coc made me feel very much a member, while respecting the requirements of my strange kind of work. I also thank the Department of Archaeology, Belmopan, Belize, for a sense of collegial affiliation during the fieldwork period. Many Mopan people do not like to be named in public. The individuals and families with whom I spent time and who shared talk with me are in any case too numerous to mention. But they will know who they are, and will allow me to remember them and to thank them once again. B'o'tik te'ex, a k'amajene' ex ichile'ex a kaj. Max k'ni a bctajcne'ex, max k'ui a t:'ajene'ex ti k'as. Max mak u loxajen, max mak u k'eyajen. Chen ki'e'exten; chen ki'ijnte,ex. PCX chen Dioos u vile'ex, ichile'ex a kuuchil, ichile'ex a naj. B'o'tik te'ex ti a f a n .
The project was facilitated and shaped in its early stages by my contacts with Ward Goodenough, Ruben Reina, and Bambi Schieftelin. John Lucy played a major
VI
ACKNOWLEDGMENTS
role in the realization of the initial manuscript, both from the point of view of practical advice and from that of theoretical refinement. Peter Danziger and Karen McCollam obliged me with discussion of quantitative matters. Any remaining faults are certainly the result of my failure to take the excellent advice of these interlocutors. Financial support for the fieldwork and analysis phases of the project was provided by the Social Sciences and Humanities Research Council of Canada (Award No. 452-87-1337), the Wenner-Gren Foundation for Anthropological Research (Grant No. 4850), and the University of Pennsylvania. The University of Virginia and the Cognitive Anthropology Research group of the Max Planck Institute for Psycholinguistics provided much-appreciated support for revision of the original manuscript. Some of the ideas presented here appeared in preliminary form in the Proceedings of the Stanford Child Language Research Forum (Danziger 1993). I thank the SCLRF for that opportunity. Gertie de Green, Carlien de Witte, and Chris Haley made substantial contributions to the material production of the text and figures. 1 come from a small family, every member of which made the effort to visit me during the period in which this investigation was conducted. The importance of the visits cannot be overestimated. Each visitor added something to my vision of the Mopan, and each visit added to my understanding of the Mopan view of family relationships. Thanks to Kurt Danziger, Mavis Waters, Ruth Danziger, Peter Danziger,and Dale Blue for making sure that the Mopan would become a part of our collective experience.
CONTENTS
Guide to Abbreviations and Orthography
ix
Chapter 1. Kinship, Semantics, and Linguistic Relativity Mopan Kinship 3 Kinship Theory: Universal and Relative 4 Linguistic Relativity 7 Psychological Reality 12 Abstract of Method 14 Chapter 2. The Mopan Setting 18 Ethnographic Setting 18 The Ethnographer's Expectations and Experiences Chapter 3. The Meanings of Kinship Terms The Piagetian Definition Task 25 The Question of Kinship 28 The Tzik Relationship Terms of Mopan Chapter 4. Tzik and Kinship 38 The Native Speaker's View 3H Mopan Life Ich Naj ('in the House') 43 T:ik Relationships and Their Socialization
25
30
47
3
23
VIII
CONTENTS
Chapter 5. Creating T;ik Relationships 53 Ritually Established Relationships 53 Acquiring Relationships through Marriage Performative Aspects of T:ik Appellation What Does Generation Mean? 64
56 61
Chapter 6. Three Semantic Analyses and Their Consequences Semantic Analyses 68 Acquisition Outcomes and Linguistic Relativity 77 Chapter 7. Formal Findings 79 Putting the Question 79 Data Tabulation 83 Data Analysis 87 Chapter 8. Language, Thought, and Reality Review 93 Some Explanatory Explorations 99 Conclusion 103 Notes
105
References Index
113 123
93
68
G U I D E TO ABBREVIATIONS AND ORTHOGRAPHY
Mopan Maya terms are rendered here in the orthography recommended by the Academia de las Lenguas Mayas de Guatemala (England and Elliott 1990). In this orthography the midcentral vowel (schwa) is designated with the symbol "a." The apostrophe signals that the preceding consonant is glottalized. Doubling of vowels indicates length. Terms of Mopan and English mentioned in the text for linguistic consideration are italicized. All Mopan forms inserted into text are also italicized. An English gloss is given for each such insertion, enclosed in single quotes. Sentence-length quotations from Mopan are set apart from the accompanying text and are italicized. They are accompanied by morphological glosses, as well as by free translation. In the gloss lines, a hyphen indicates a morpheme boundary and an underscore connects the several English glosses applicable to a single "portmanteau" morpheme in Mopan. In a few cases involving Mopan formal speech genres, paragraph-length citations have been provided with free translation only. These are intended as illustration of the content of such speech only. Full presentation of these texts, with adequate glossing and English rendition, would be matter for a different book. Abbreviations tor these morphological glosses appear below. For further grammatical detail, the reader is referred to Ulrich, Ulrich, and Peck (1987) and to Danziger (1994, 1996b): 1 = first person 2 = second person 3 = third person A = actor pronoun (transitive actor, active intransitive participant, possessor).
ix
x
GUIDE TO ABBREVIATIONS AND ORTHOGRAPHY
APP = applicative B = undergoer pronoun (transitive undergoer, intransitive participant) CL = numeral classifier CMPL = complementizer COMPL = completive aspect DAT = dative DET = determiner DUB = dubitative DUR = durative DX = deictic EMPH = emphatic particle HAB = habitual HUM = humiliative I = irrealis INC = incompletive aspect INCH = inchoative IND = independent pronoun NEG = negative particle NOM = nominalizer OBLIG = obligative PERF = perfect aspect (archaic in Mopan) PL = plural PREP = preposition Q = interrogative/conditional particle QUANT = count scope specifier SCOPE = scope specifier SUBJ = subjunctive TR = transitivizer
Relatively Speaking
This page intentionally left blank
1
KINSHIP, S E M A N T I C S , AND LINGUISTIC RELATIVITY
This book investigates the principle ot linguistic relativity by exploring the cultural
and psychological reality of three different models of kinship conceptualization among speakers of Mopan Maya, an indigenous language of Central America. I ask whether it is the case that across languages and cultures, individuals find certain concepts natural in experience, implying that where languages vary in their expression of these concepts they do so only superficially. Or, instead, does the structural organization of languages actually help to form the categories of thought which afterward appear most natural to their speakers? Mopan Kinship
Every language in the world has words for family members (see figure 1.1). But the exact range of meaning of these words varies enormously across languages and cultures. In Mopan Maya, for example (see figure 1.2), the word that is used for one's older brother (suku'un) is also used for a parent's younger brother.1 A different word (tataa') is used for a parent's older brother, and that same word (tataa') is also used for what we would call a grandfather or a great-uncle. Meanwhile, brothers and sisters younger than the speaker are called by a different term (itz'iin). It is certainly true that if a Mopan individual points to a particular family member—say, an elder male sibling—an English speaker can find a word to express the relationship in question (in this case, the English word brother). But does the English word brother mean the same thing as the Mopan word suku'un, simply because 3
4
RELATIVELY
SPEAKING
Figure 1.1
Terms for some English family relationships
the two words overlap in this (important) portion of their denotation? Or do the radical differences in the range of reference between these two words actually make a difference to the ways that Mopan and English speakers conceptualize this "same" family relationship? Kinship Theory: Universal and Relative Observations like the Mopan ones have given rise to generations of debate within anthropology (Schneider 1965, 1984; Scheffler and Lounsbury 1971; Goodenough 1956, 1965-Xeach 1962; Needham 1962 [I960], 1971; Lounsbury 1956, 1965, 1969a; Murdock 1965 [1949]; Kroeber 1909, Morgan 1970 [1870]). The questions that have arisen have dealt with the universality of genealogy as a basis for social organization worldwide. At another level, however, these classic debates over the place of genealogy in kinship posed, in the arena of linguistic semantics, the time-honored questions of linguistic relativity. The specific question that has vexed the field is this: Is the semantic analyst justified in proposing that certain aspects of a single word's meaning are more impor-
Kinship, Semantics, and Linguistic Relativity
Figure 1.2
5
Terms for some Mopan family relationships
taut, more central than others? If we could do this, for example, to solve the Mopan puzzle above, we could relegate those aspects of the meaning of Mopan siiku'un that do not overlap with the meanings of English brother to a conceptually peripheral or secondary status, thus proposing that the Mopan and the English words do indeed mean essentially the same thing (cf. Lounsbury 1964, 1969a). We could suggest, for example, that the central referent of Mopan suku'un is actually 'male sibling', by virtue of the natural primacy of the nuclear family over more distant genealogical bonds. We know that the word suku'un is also used, say, for certain of the parent's siblings, but, in this view, that is purely by extension from the more central and more natural nuclear family referent.2 If we take this approach in the analysis of languages and cultures other than our own, however, we necessarily prejudge the question of the primacy of the nuclear
6
RELATIVELY SPEAKING
family in the society in question. Also and more generally, we prejudge the question of the possible influence of cultural and linguistic structures themselves upon conceptualization. Most famously, Benjamin Lee Whorf (1956 [1939]) has argued that linguistic classifications are psychological entities in their own right, and that to understand them as such is an excellent way to broaden our own conceptual horizons. If he was right, then the Mopan category siiku'un should constitute an undivided and culturally unique conceptual unit, the meaning of which would have to be discovered ethnographically, and the existence of which would raise serious doubts about a universal and familiar domain of genealogically calculated kinship across all cultures (Needham 1971; Leach 1962). To most cultural anthropologists engaged in theorizing kinship today, the latter interpretation appears not only intellectually obvious but politically compelling (Peletz 1995; Schneider 1984). The traditional categories of kinship in anthropology have been thoroughly deconstructed and are today understood to be symbolic and cultural creations rather than natural necessities. Even such apparently fundamental primitives of genealogy as 'man', 'woman', or 'mother' are understood to be socially performed and culturally constructed, rather than naturally given (Morris 1995; Strathern 1992). In view of these developments in the anthropology of kinship, it is interesting to note, on the contrary, that a view of nuclear family-based kinship as universal and natural is very much alive in current cognitively oriented attempts to discover various universal domains of modular and perhaps innate human knowledge (Wierzbicka 1992; Keesing 1990; Hirschfeld 1989). In addition, despite its controversial status within anthropology, Lounsbury's claim that the nuclear family is conceptually primary for all speakers regardless of their local terminological system is cited as foundational in the new fields of prototype semantics and cognitive linguistics which take semantic organization in terms of central and peripheral referents as their guiding principle (Kronenfeld 1996; Lakoff 1987, 22-24; see also D'Andrade 1995, 105). Despite the existence of such radically different views, both of cultural variation in the kinship domain and more generally of the nature of human thinking in crosslinguistically variable arenas, little current debate addresses these issues directly. Indeed, opinions about what constitutes adequate evidence and about what convinces in argumentation have diverged so radically across culturally and cognitively oriented studies of meaning that today such discussion sometimes appears almost impossible (cf. Bruner 1990, Hirschfeld 1986). In the present study I attempt to bridge this discourse gap. I bring the culturally informed perspective of the older anthropological debate into the modern cognitively oriented discussion of words and their meanings. Aligning classic debates over kinship semantics with the current renewal of interest in the measurable effects of language variation on thought (Levinson 1998; Pederson et al. 1998; Gumperz and Levinson 1996; Lucy 1992a), I take up the challenge of determining the psychological reality of alternative semantic models of kinship vocabulary. I endeavor to do so in a manner that can be heard across the different disciplines concerned, by carrying out an investigation that integrates ethnographic, linguistic, and psychological observations to discriminate between competing models of kinship conceptualization in speakers of the Mopan Maya language. Since the different models proposed also make different claims about the universality of human conceptualization
Kinship, Semantics, and Linguistic Relativity
1
under cross-linguistic variation in expression, the investigation is also one that distinguishes among different positions with respect to linguistic relativity.
Linguistic Relativity As propounded by Benjamin Lee Whorf (1956c [1940], 214), the principle of linguistic relativity claims that an individual's unreflective acts of conceptual classification (''habitual thought"; Whorf 1956b [1939]) will follow the same patterns that also guide conventional categorization in his or her language (see Lucy 1997a, 1992b for extensive recent review). This entails that one's view of reality is linked in an extremely significant manner to the patterning of one's native language, and that cultural worldviews will vary around the globe as languages and dialects vary. Rather than labeling preexisting or universal concepts, linguistic structure in this view provides much of the architecture from which seemingly natural interpretations of reality are constructed. This in turn, according to Whorf, means that the study of languages is among the most liberating of intellectual exercises for those who would discover alternative and unexpected perspectives on the assumptions according to which they live their own lives (Whorf 1956c [1940]). Monosemy The principle of linguistic relativity is built solidly upon the notion that one of the most important kinds of linguistic meaning is that which is carried in structural contrast (Saussure 1959 [1916]). Sapir (1925, 1949) pointed out that where a principle of structural contrast (e.g., an opposition in phonology between voiced and voiceless consonants) is used again and again in a language and under inflection for different values, this dimension of contrast itself becomes psychologically real to speakers of the language—but not to those whose language does not employ such a contrast. In the area of speech sounds, Sapir's idea has been fully vindicated since his first writings on the subject (Jusczyk 1996). People actually do have different perceptual intuitions about the speech sounds they hear, depending on the phonological organization of their native language. Meanwhile, in direct adaptation of this idea to the conceptual plane, Sapir (1921) and later Whorf (1956b [1939]) suggested that repeated conceptual contrasts (e.g., a semantic opposition made in some languages but not in others between mass and count nouns—but also and more relevant to kinship studies an opposition made in some lexical paradigms between male and female, lineal and collateral or junior and senior) would result in exactly these conceptual dimensions of contrast becoming psychologically real (i.e., the source of subjectively convincing intuitions about the conceptual naturalness of the contrast in question) to speakers of that language but not to speakers of languages in which the contrast was not made. In one example, Whorf (1956c [1940]) suggested that the fact that speakers of European languages philosophize in terms of an intuitive distinction between objects and events can be traced to the fact that European languages make a grammatical distinction between nouns and verbs. Citing such dubious natural objects as the ones represented by
8
RELATIVELY
SPEAKING
English nouns like spark and fiat, and such anomalous natural events as those represented by English adhere or dwell, Whorf argued that speakers of a language that did not make a noun-verb distinction in grammar would not find the conceptual distinction between objects and events self-evident either.3 Whorf suggested that despite the fact that languages often group together potentially quite disjunct sorts of referents, speakers in general prefer to find or construct a single conceptual meaning for each formal structure of their own language. The single common denominator is then imbued in speakers' reflections with the qualities of an intuitively natural reality. For Whorf, this preference on the part of ordinary speakers is a strong one, and it operates without regard to what experts might know about the history or typology of the linguistic grouping in question. In the terminology of today's linguistic semantics (see Ruhl 1989), Whorf claimed that, everything else being equal, speakers assume conceptual monosemy—a relationship of only one meaning to every linguistic form—to hold as the default in their language (Croft 1993; see also Lucy 1992b, 47 for a general exposition of "Whorf's basic argument" in terms of disjunct reference). In one illustration (Whorf 1956a [1941J, 261-62), Whorf describes how a woman of his acquaintance owned a type of cat known conventionally as a "Coon cat." She was, reports Whorf, convinced that her cat was related to the raccoon species, and in fact saw aspects of physical resemblance between her pet and the raccoons. In point of historical fact, the cat type is evidently named for a certain Captain Coon, but this was unknown to the woman in question. Even when the experts endeavored to enlighten her, she continued to reject their diachronic perspective, and to insist that her cat indeed possessed a very unusual cross-species pedigree. In fact, this woman understood the phonologically identical forms coon (in the popular varietal name of her cat) and coon (as a shortened label for the raccoon species) as also semantically identical. For Whorf, the point of the anecdote (which is clearly best taken as didactic illustration rather than as empirical demonstration) is to show how a cognitive insistence on monosemy is the mechanism which powers the psychological effects he claims in the linguistic relativity phenomenon. In another, more complex, but much more far-reaching example, Whorf argued that beyond the facts of our language, there are plenty of experiental reasons to believe that the referents of the English words day and man are conceptually very different from one another. But in English, both these words can be pluralized. Whorf argued that this linguistic fact has been enough to lead speakers of English to assume that there is something conceptually in common to the referents of day and of man—specifically the (ascribed) semantic attribute of discrete individuality that licenses the application of the English plural morpheme. If day were not a count noun in English, says Whorf, as its translation equivalent is not in Hopi, we might not so habitually conceptualize each twenty-four-hour cycle as a new individual entity. 4 For extensive further discussion of this example, see Lucy 19920a and 1992b. Polysemy Where it can be applied, an assumption of monosemy is not only, as Whorf proposed, the handmaiden of relativity; it also has clear practical advantages. To the field
Kinship, Semantics, and Linguistic Relativity
9
linguist it represents Occam's razor (Sasse 1993). To language-learning children (Bowerman and Choi 1991; Bowerman 1997) and, indeed, to ordinary people in everyday thinking (Bruner, Goodnow, and Austin 1956; Kronenfeld 1996), it represents cognitive efficiency. It has also long been recognized at the level of linguistic theory, however, that to analyze semantic classes always in terms of synchronic and language-specific monosemy poses serious problems for both typological and historical study (Croft 1999; Lounsbury 1969a). And, most recently, the claim has been increasingly made (Lakoff 1987; Langacker 1986, 1987; Kronenfeld 1996; Palmer 1996; Taylor 1989) that not only linguistic analysts but also ordinary speakers conceptualize the categories of their language not in terms of monosemy and contrastive structure but as groupings of multiple and differently weighted referents, of which one is clearly central with respect to the others. The weightings involved, and speakers' intuitions regarding the identity of the central referent, derive from nonlinguistic experience. Such a model of conceptual understanding is linked to semantic analyses in terms of polysemy in linguistic semantics.5 The immediate ancestry of the recent heightened interest in polysemy is generally acknowledged to lie in the anthropological literature of enthoscience, including kinship semantics. Most often cited (Lakoff 1987, 22-24; Palmer 1996; D'Andrade 1995; Kronenfeld 1996) are Berlin and Kay's (1969) study of color terminologies, and Lounsbury's (1964, 1969a) work on the focal referents of kinship categories.6 These authors couched their work explicitly as a response to linguistic relativity as expounded by Whorf. In these early studies, the strongest rhetoric arose from the claim that speakers across languages would agree with one another on the identity of focal referents, regardless of the specific lexical segmentation of the domain that might be imposed by their particular language. This claim was presented forcefully as an explicit contradiction to the Whorfian proposal that differences in language categories across cultures could result in—or even show parallels in—-differences in conceptualization. Lounsbury's (1965, 1969a) motivation in proposing universal focal referents for kinship categories was a typological one, based on the desire to preserve a common ground for comparative inquiry. At the time he was writing, the comparative domain of kinship as a cross-culturally universal domain based on genealogical calculation was under siege in cultural anthropology. The attack came from cultural relativists who proposed (monosemously) that unfamiliar cultural categories should be taken at face value, even if this meant that the comparative domain would cease to exist (Leach 1962; Fortes 1962; Needham 1971). Lounsbury's move in this debate was to propose that in order to save kinship as a comparative domain it was necessary only to postulate that that referent of any kinship term which was genealogically closest to Ego should count universally as the focal referent. Specific extension rules could then derive other referents from this focal referent in different ways in different languages. He argued that his cross-language identification of the same central referent across languages was confirmed by the fact that the number of types of kinship terminology systems found worldwide was relatively small (see also Murdock 1965 [1949]), and that each of these very few types could be understood as the outcome of applying a particular set of extension rules to the same set of central referents. 7
10
RELATIVELY SPEAKIMG
Today, however, few advocates of general polysemy in semantics would defend the notion that focal referents must be universal across languages, or in fact that this is the main point of doing semantics in terms of polysemy (although see Croft 1999). In the wake of kinship and color, the method for identifying central referents in semantics was quickly converted from the demonstration of cross-language to that of within-language agreement about the identity of best example ("prototype") members of named categories (Rosch 1978). In short order, within-language speaker agreement on the central members of clearly nonuniversal lexical categories such as furniture (Rosch 1978), speech act verbs (Coleman and Kay 1981), and indeed even numbers (Armstrong, Gleitman, and Gleitman, 1983) was demonstrated.Today 's practitioners of polysemy in semantic analysis are usually content to leave cross-cultural issues implicitly to one side or to propose culturally particular focal referents for the categories they discuss (even in the domain of kinship; see Kronenfeld 1996). In the present context it is important to be clear, however, that while a semantic analysis in terms of central referents need not claim universality across cultures, and may even be compatible with the claim that conceptualization differs across cultures, it cannot in all logic be compatible with a strictly linguistic relativity, even in the language to which it does apply. The conceptual primacy of the central referent under polysemy, in fact, exists in defiance of overt language structure, which on the contrary shows various other referents to be equally designated by the same term. Under analyses based on polysemy, the basic claim must be that psychological apprehension of meaning in linguistic categories depends not on linguistic structure but on something else. What that something is has been variously construed by different theorists. As in the original proposals from color and kinship, it may be referred to universal propensities of the human physiology or psyche. Under nonuniversalist accounts, it may be specified as related to functional salience or frequency of encounter with the central referent (Hunn 1976, 1985). It may be referred to the structure of the body (Lakoff 1987), the world (Rosch 1978), or the mind (Rosch 1988), or it may simply remain unexplained. Regardless of the mechanism invoked, however, and although analyses made in terms of polysemy may be very sensitive to issues of cultural variation, such analyses by their very nature must deny primary psychologically constitutive power to linguistic classification itself. In his classic studies claiming polysemous organization of kinship categories, Lounsbury proposes extension rules that are very deliberately stated so that they operate over strictly genealogical categories.As we have seen, for Lounsbury, this was the rhetorical point of focal referent analysis. But it was also clear to Lounsbury that logical equivalents for these extensions could be found to exist, in every case, in the culture-specific sociological correlates of each extension rule (1969; see also Goodenough n.d.). In one much-debated example, a plausible sociological motivation for the Crow-type generational skewing of the Trobriand Islands was offered by Leach (1962), who appealed to the details of an economy governed by matrilineal descent but existing in the context of virilocal residence. Leach claimed that the complex patterns of economic and affective association that result from this combination of social circumstances, for any individual (male) Trobriander, would result in the functional clustering of just those groupings of relatives that, if
Kinship, Semantics, and Linguistic Relativity
11
genealogically defined, show the distinctive Crow-Omaha pattern of generational skewing. But if such logical equivalence is a possibility, why not reverse the psychological analysis and cite these motivating but nongenealogical generalizations as the very conceptual principles that define the domain? This was precisely the move being advocated by Lounsbury's relativist colleagues, who claimed that the "kinship" categories of the Trobrianders just cited, for example, were not the products of genealogical calculation from universal central referents but instead were sociologically defined at the level of the total category. For Lounsbury, such a course was unacceptable because it involved abandoning genealogical kinship as a comparative domain—in no two cultural cases would the sociological correlates of genealogical extension rules be exactly alike. But for most semanticists who today favor polysemy as the default analytical assumption, we have seen that a concern to maintain universal domains of comparison is not an especial priority. For these, the logical interconvertibility of extension rules with defining principles of a category can be a strength rather than a weakness. This fact allows some current theorists to claim that semantic analyses made in terms of central referents can also support linguistic relativity (cf. Lakoff and Johnson 1980, Lakoff 1987, 334-3V).8 To ask then, whether we might find a place for linguistic relativity in a semantics based on polysemy is to ask what, if any, is the psychological weight to be assigned to the semantic extensions that are drawn from the proposed central members of language categories. And what, by contrast, is the weight of the (nonlinguistically derived) central referents? For the puiposes of linguistic relativity, to the extent that polysemous analysis focuses upon conceptualizations and actions that follow from the extensions drawn from the central referents it proposes, it may approach the relativist possibilities of monosemy. To the extent that polysemous analysis focuses instead on conceptualizations and actions based on centrality of focal referents, it diverges from the tenets of true linguistic relativity. For the purposes of the present exploration, "polysemy" is taken to claim psychological primacy for the focal referent and to assign relatively less psychological weight to the extensional links which unite focal and peripheral referents. Homonymy: A Third Logical Possibility In the tradition of kinship semantics and elsewhere, the monosemy-polysemy debate arises to solve the problem of disjunct reference: the case in which a single linguistic item appears to the analyst to apply to more than one potential class of referents. In addition to these two options, however, a third solution is also widely recognized. This is the strategy that refrains from attempting to solve the puzzle of disjunct reference and that instead allows the different sets of referents to remain disjunct in the analysis. This is an analysis in terms of homonymv (in the vocabulary of the philosophy of language, an extensional view of meaning, in which the different referents of a term are represented as an unordered list). 9 Such an analysis proposes that, as under polysemy, the same linguistic form covers multiple psychological meanings, but it does not propose that these are related to one another in the understanding of speakers. It is possible, for example, to solve my opening puzzle by proposing that the
12
RELATIVELY SPEAKING
Mopan word sukii'iiH means 'elder brother', and that it equally means 'youngish uncle', but that no conceptual relationship between these two meanings exists for Mopan speakers. In the same way, a homonymous reading of English brother could argue that brother means both 'elder' and 'younger' male siblings, while claiming that no conceptual relationship between these two senses of the word exists for English speakers. While homonymy does not perhaps appear intuitively appealing in this case, there are many cases in every language in which it is a compelling analytical option (e.g., in English, when the word bank is used to refer both to a financial institution and to the edge of a river). The possibility of semantic homonymy will be fully respected in this book. Psychological Reality Systems of kinship terminology are notorious within anthropology for the multiplicity of semantic models that can be drawn up to account for them (Burling 1964). It has also long been recognized in kinship studies that although formal analyses might proliferate, there was no way to tell which analysis might correspond to the way in which users of a language actually conceptualize the meanings of terms (Wallace and Atkins 1960). For that, other measures of psychological reality would have to be used. The call for independent psychological testing of semantic analyses has also been sounded more recently as models based upon the assumption of polysemy have multiplied within the new discipline of cognitive linguistics (Sandra and Rice 1995; Croft 1998). This book offers an empirical test of the psychological reality of competing models of semantic organization in the domain of family relationships in Mopan Maya. I set out to define the universe of interest for Mopan in terms of a particular speech act (that of respectful greeting), which characterizes one kind of situational encounter between Mopan Maya individuals. This approach has a formal motivation, in that it allows me to isolate a set of terms for study without presupposing critical aspects of cultural meaning—in particular, the prior existence of a domain of kinship. It also, however, involves the project in detailed consideration of a peculiarly Mopan set of social relationships. I take up the ethnographic description of the kinds of cultural relationship recognized in the Mopan "respect" greeting, and the ways that distinctions among such relationships are recognized and manipulated. I illuminate—especially from the vantage point of Mopan women—a particular instantiation of the pan-Mayan cultural notion of 'respect'. Observable Reflexes in Reflective and Habitual Thought Because they are concerned with abstract principles or features that characterize all referents rather than with any referent in particular, analyses made in terms of monosemy differ critically from those made in terms of some central referent in locating the essentials of conceptual meaning at the boundaries of lexical categories. At exactly those points, most distant from any typological or intuitive referential center, one lexical or grammatical category is distinguished from its neighbor in a
Kinship, Semantics, and Linguistic Relativity
13
particular language. Where semantic categories are organized in terms of monosemy, we expect that the boundaries of the categories will be clear, and that judgments of whether particular entities are or are not examples of the category in question will be categorical and untroubled. By contrast, semantic analyses in terms of polysemy locate the essentials of conceptual meaning not at the boundaries of linguistic categories but at their intuitive or typological centers. Under this kind of analysis, the boundaries of linguistic categories are held to be conceptually "fuzzy"; speakers within a language community are expected to hesitate or disagree when asked to judge whether peripheral referents are properly members of a given linguistic category—while at the same time they will agree wholeheartedly that others, more central, definitely belong (Coleman and Kay 1981; Rosch 1978). The canonical method for demonstrating the existence of semantic organization in terms of polysemy has been established as a procedure of decontextualized reflective introspection, whether this is done singly on the part of the analyst alone as in much of cognitive linguistics, or concurrently by multiple introspecting consultants as in Rosen's original studies. Reflective judgments as to the identity and existence of best examples, and also as to the fuzziness of referent membership at category boundaries (is a potato a vegetable'!), are the accepted diagnostics of this type of semantic organization. Quite on the contrary, evidence for the kinds of psychological effects predicted under linguistic relativity is expected to be observable at the level of habitual or nonreflective thought and behavior, and at category boundaries. The existence of semantic polysemy in the form of a reflective willingness to state a best example therefore does not vitiate the possibility that in nonreflective thought, linguistic categories could also show Whorfian effects. 10 It is possible that prototype effects are those that emerge from decontextualized reflection while category-boundary effects are those that operate in everyday nonreflective and contextualized functioning. We can therefore allow ourselves to ask Mopan consultants for reflective identification of central referents of their kinship terms, and at the same time ask in other, less self-conscious or more situated ways, whether language-specific category boundaries in the same domain also have consequences for habitual thought and behavior. In the years since the heyday of studies in kinship semantics and of prototype effects, it has become increasingly clear that the act of linguistic classification in any particular case depends on what one is classifying for (Levinson 1983, Frake 1980). A tissue of culturally defined felicity conditions consisting among other things of discourse conventions, cultural expectations, and personal motivations provides an interface between the word's meaning and its contextual interpretation, and links every utterance, by its function, to the speech situation in which it occurs (Goffman 1983). In turn, the situation in which any utterance occurs is always modified by the occurrence of the utterance in question. In this book I undertake to present, in addition to the reflective commentary of Mopan speakers regarding the best examples of referents for their kinship terms, two kinds of evidence also relevant to the case for linguistic relativity. On the one hand, a formal test in psycholinguistic style is designed and implemented to access habitual rather than reflective thought. On the other, and in the course of implementing the
14
RELATIVELY SPEAKING
formal test, I use the anthropological methods of ethnographic observation and involvement to offer an extended description of the Mopan domain itself, and of its cultural constitution in context. It will become apparent that although the boundaries of the Mopan domain, as well as the boundaries between terms within the domain, appear unclear under an analysis made out of context, they are not permitted to remain so in social practice. Combining quantitative with qualitative methodology, this work thus draws out and examines the vital links between social puipose, lexical categorization, and cognitive organization. At one level, it remains an investigation of word meaning, which uses the notions of linguistic context and function to facilitate an exploration into theoretical semantics and which contributes to our empirical knowledge of the phenomenon of linguistic relativity. At another, it inevitably becomes an investigation of the Mopan context itself.
Abstract of Method Fourteen months of ethnographic fieldwork in southern Belize provide the basis for an understanding of the Mopan kinship domain. Use of family relationship terms in obligatory greeting is consistently understood by Mopan as an important instantiation of religious respect (Mopan tzik). A deep reverence for age as the vehicle of knowledge parallels the attention to the factor of relative age, which, as we have seen, is crucial to Mopan kin term semantics. But relative age is not always understood by Mopan to be predicated upon objective factors such as biological generation or absolute chronology. Rather, the distinction turns on collective assessment of the various social and cultural criteria that allow one person to play a teaching and nurturing role with respect to another. Also highlighting the cultural rather than the natural bases for classification within this domain, Mopan kin relationships are regularly established and maintained through performative speech acts (i.e., 'baptisms', greetings, and poetic conversational exchanges). Nevertheless, according to Mopan, relationships within this domain are ultimately based upon parent-child links, and nuclear family referents are identifiable within the Mopan categories. A more universalist semantic analysis in terms of central referents is therefore quite feasible, as suggested earlier. On the basis of the ethnographic description, three very different models of kinship semantics are proposed for Mopan. One, an approach in terms of monosemy, proposes that a peculiarly Mopan feature of fitness for nurture by virtue of maturity functions in Mopan cognitive organization to define the category as a uniquely Mopan conceptual whole. This analysis proposes that category boundaries in the Mopan domain are well defined and psychologically salient rather than fuzzy and peripheral. The second, an analysis assuming polysemy, and made in terms of central category members, follows up linguistic and interview evidence that indicates the reflective existence of familiar typical referents within the nuclear family for each Mopan term. A third analysis assumes homonymy and uses ethnogenealogical data to make an analysis in terms of an unordered list of relative products of genealogical primitives.
Kinship, Semantics, and Linguistic Relativity
15
Next I take up the challenge of establishing the cognitive status of these three possible analyses. Previous cross-linguistic investigations have observed that the semantic makeup of different kinship terms within one language conditions the order in which their relational aspect is grasped by children (Haviland and Clark 1974; Piaget 1928). On this basis, I am able to predict that children's Piagetian acquisition of selected Mopan kinship terms will proceed in three alternative sequences under the three competing semantic models (cf. Greenfield and Childs 1977). The null hypothesis pi-edicts that the sequence of acquisition will correspond to chance. Children's Language The method adopted in the formal study observes Mopan children's developing metalinguistic abilities and interprets these in the light of previously documented cross-linguistic developmental regularities related to semantic structure. In this way, it becomes possible to distinguish among the three competing types of semantic model. Ontogenetic data are used to adjudicate between different theoretical positions that had originally been identified on the basis of observations among adults. Inversely, however, even as it uses acquisition data to distinguish among alternative models of adult semantic organization, this book also constitutes an inquiry into the growth of Mopan children's understanding of one area of the lexicon in their language and contributes more generally to the global project of the cross-linguistic study of language acquisition and of cultural learning. Accompanying this aspect of the formal study, the ethnographic portions of the book pay special attention to describing the cultural contexts and social situations in which Mopan children encounter and use family relationship terms and concepts. Although my ultimate intention is broadly comparative, and although Euro-American data and hypotheses provide a framework, I compare a given child's acquisition of certain terms only with his or ^er own acquisition of other terms, or else with the acquisition of the same term by otper children of his or her own culture and language. T n language acquisition studies today, nativists and nonnativists alike often view language categories as superimposed onto prior and universal conceptual categories derived either from genetics (Chomsky 1988) or from universal experience based on physiology and action (Lakoff 1991; Slobin 1985). Embedded in this majority position is a view that infants approach the world of experience—including linguistic and cultural experience—as individuals, and in isolation. The acquisition problem is seen as one of matching words to things, using metaphorical motivation if that is necessary. But a minority of specialists raise the issue of language itself as possibly instrumental in the formation of concepts. The language socialization paradigm developed by Schieffelin and Ochs (1986), for example, assumes an early sensitivity to the activities of nearby humans as a primary source of input data for the developing child and sees "experience" as other than merely individual and physiological. Bowerman (1985) also argues powerfully that the hypothesis of the existence of semantic concepts prior to language is susceptible to disproof, through comparative documentation of the course of acquisition in specific domains of different languages. A language acquisition position sympathetic to linguistic relativity argues, therefore, tht't certain kinds of cognitive prin-
16
RELATIVELY SPEAKING
ciples (metaphor, motivation) may themselves be brought into play by the prior existence of linguistic classification as a social fact. In the arenas of both adult cognition and child language, then, the ultimate philosophical questions are similar. Are linquistic meanings given by the structure of the world, the mind, or of individual experience, or do they exist instead as culturally and linguistically framed concepts, often defined in relation to one another rather than in any discrete and objective sense? In achieving results that distinguish between these viewpoints in theoretical semantics, I also achieve results that bear upon general issues of language learning in cultural context. Two Findings In the Mopan study, data from one hundred children aged seven to fourteen years provide results that most strongly support the psychological reality of the analysis which assumes monosemy. The findings are therefore Whorfian in the traditional sense: even while speakers' reflective intuitions show best-example members of lexical categories, linguistic categorization itself is found to correspond to an aspect of nonreflective psychological reality. But in its conclusions, the study goes one step further. It also examines the social and discourse mechanisms by which Mopan linguistic categorization in the kinship domain could come to have such psychological force. Mopan ethnography shows us that in the situations to which 'kinship/respect' (tzik) is relevant, clear-cut and mutually exclusive courses of linguistic action (i.e., whether or not one performs socially constitutive acts like baptism and greeting) have well-defined and important social consequences. The local circumstances therefore do not allow for gradations of membership in tzik -related categories. Kinship terms themselves are found in turn to have the relativity-relevant property that their use readily creates, as well as merely representing particular kinds of sociality. The study thus concludes that while language-specific structures may indeed have psychological reality, this fact itself may be culturally relative and subject to the mediating agency of particular social and communicative practices. Plan of the Book In chapter 2,1 introduce the Mopan Maya and offer a general framework for the ethnographic description that appears in later chapters, through discussion of the previously documented cultural notion of 'respect'. In chapter 3, the specific method for the formal study is introduced, with an explanation of how it can distinguish among alternative semantic models of adult cognition. Informed by this discussion and by ethnographic exploration of the Mopan concept of tzik 'respect', the problem of isolating a Mopan domain for comparative study is solved through explication of the dependence of the selected task on certain semantic and pragmatic linguistic properties of kinship terms. That is, I refrain from presupposing either the content of the domain or the meaning of the terms within it. I do not seek Mopan kinship terms. Rather, I commit myself to inclusion in the study all of the terms that occur at the intersection of two defining frames—one formal (semantic), the other based on usage (pragmatic).
Kinship, Semantics, and Linguistic Relativity
17
Chapters 4 and 5 are almost entirely ethnographic. They discuss the kinds of social relationships to which 'respect' relationship terms apply and the ways in which Mopan children are exposed to and experience these types of relationships. An effort is made to understand the cultural ways in which such relationships are distinguished from one another. This stage of the research involves requests for Mopan reflection on the meanings of their terms, as well as careful ethnographic observation and interpretation of the kinds of Mopan contexts in which these terms are used. Such an approach necessarily involves consideration of the peculiarly Mopan nature of the social relationships under discussion. I discover in particular that a notion of relative seniority, which has already appeared in the brief discussion here of the semantics of Mopan suku'itn, has wide repercussions in Mopan society. It is related to cultural values that privilege tradition and custom rather than innovation and novelty, and which rely on personal testimony and individual leaching rather than on impersonal or public media for dissemination. Of particular interest here is the fact that respect greeting relationships may be performatively established through negotiated interaction in culturally appropriate situations. In chapter 6, three alternative formal analyses, respectively assuming monosemy, polysemy, and homonymy in the Mopan domain of respect-greeting relationships, are presented. Mutually exclusive predictions for Mopan children's performance on the Piagetian task under the three analyses are made. Chapter 7 presents the specifics of task administration and the results of the formal study. These include a preliminary replication, for the Mopan, of others' cross-cultural findings in the same task. The final results support a monosemous analysis of the Mopan domain. In chapter 8, this result is integrated with a review of my ethnographic findings to inform some more general conclusions about the relationship of culture to language, and of both to habitual conceptualization.
2
T H E MOPAIN S E T T I N G
Ethnographic Setting Mopan is a member of the Yucatecan group of Mayan languages and is spoken by several thousand people in the Peten regions of Belize and Guatemala. Although the language is mutually intelligible with the Yucatec, Itzaj, and Lacandon languages of the Mexican peninsula, it exhibits significant phonological, lexical, and structural differences from them (Danziger 1994, 1996b; Ulrich, Ulrich, and Peck 1986; Ulrich and Ulrich 1976, 1966; Thompson 1930). The Mopan language in Belize features many Spanish lexemes, borrowed during an earlier era when that language was spoken in Guatemala (cf. Verbeeck 1998). Today, however, Belizean Mopan in general do not speak Spanish and they understand these Spanish expressions as Maaya. From October 1988 to September 1989,1 lived in Belize's oldest Mopan village. I made shorter visits in 1992, 1993, and 1996. The oldest Belizean Mopan village is also the largest in a region that is home to perhaps two or three thousand speakers. Many of these people inhabit smaller villages in the area, some of which (unlike the larger village) boast a mixture of Ke'kchi and Mopan Maya inhabitants (Davidson 1987).' This oldest village is regarded by its own inhabitants (many of whom can remember the founding of the other villages) as the seat of tradition and cultural orthodoxy in the area. For all the Mopan of the area, to be in this largest village is to be ich kaj, 'in the village', the existence of numerous other villages roundabout notwithstanding. In the late 1980s and early 1990s, the village consisted of about thirteen hundred Mopan Maya speakers. 18
The Mopan Setting
19
In this village, Mopan Maya is acquired as a first language by all children of Mopan parents, and children are monolingual until they enter the school system, usually between the ages of five and ten years. A Caribbean variety of English is presented to children in the schools, and some level of facility in this variety has been achieved by many young adults and older children. Adult men in general have a better command of English than do women, but Mopan remains the language of choice in the home and in village life generally. My ethnographic data consist of the notes on everyday life and on language use that I made in the course of participant observation. I conducted formal and informal ethnographic, ethnogenealogical, and linguistic interviews. I made tape recordings of traditional stories, ritual speech exchanges, and stories from life. I scrutinized my own reactions to the spontaneous teaching of Mopan people. From the beginning, I worked as much as possible in Mopan Maya, and my competence in this language improved steadily, to the point where I was able to engage in fruitful learning and teaching exchanges with the oldest generation, most of whom are monolingual in Mopan. The Mopan understand well the idea of learning under guided tutelage, and they freely welcomed me to study their language and their way of life. They were forthcoming and even expansive on certain topics (including that of tzik 'respect' as discussed later) but were more guarded about other kinds of information, especially about what is perceived to be efficacious knowledge that may be applied by professionals on behalf of nonprofessionals for a fee. Into this second category came, for example, specialized knowledge of sickness and healing, and the information and charms for promoting agricultural increase and for placating wrathful deities. A studiedly noncommittal attitude is characteristic of the Mopan when they are presented with direct questions about these kinds of topics. Individuals will often provide answers of the type "It's up to you," "You know best," or "Only God knows," when asked directly for opinions and for certain kinds of information. Mopan Life Ich Kaj Today's Belizean Mopan are slash-and-burn cultivators of maize and beans for subsistence. They also plant rice, which they sell for cash in the nearby non-Mayan market town. They are the descendants of late-nineteenth-century immigrants from the Peten, themselves descended from colonial period Yucatecans of the region—perhaps originally even from central Belize (Thompson 1988, 20). The nineteenth-century immigrants left Guatemala in flight from compulsory military service and debt peonage, and their arrival in Belize approximately coincided with the arrival of Ke'kchi and "Kekchi-Chol" (Thompson 1930) peoples from the vicinity of Coban to the south. In Belize, the immigrant Maya established a way of life based on subsistence farming and a syncretic version of Catholicism that is similar in its broad outlines to the general Mesoamerican and Lowland Mayan pattern (Nash 1967; Redfield and Villa Rojas 1934). In the Mopan village, the ideal norm is for virilocal nuclear family households to cluster in groups constituted of a set of brothers and their wives. The physical labor of extracting a livelihood from the forest is an ever-present fact of life. Division of
20
RELATIVELY SPEAKING
labor within the subsistence economy is based on age and on sex. Men work in the fields, occupied with primary agricultural production, while women work in the village, processing raw agricultural products into food. Children work diligently and willingly alongside adults in whatever capacity they can. The very old also contribute to the household economy to the extent of their ability, choosing for themselves more sedentary and less strenuous tasks. When a particularly large task, such as building a house, planting a field of corn, or baptizing a baby must be accomplished, the Mopan traditionally proceed by means of labor exchange. On such occasions, persons standing in relationships of solidarity with members of the original household—by virtue sometimes of personal friendship, but more often also by virtue of perceived family relationship—will provide labor to the benefit of the original household and will acquire the scrupulously honored right to call on that household in turn when their own need arises. In a more recent alternative to this pattern, an individual may today hire assistants by the day for a cash wage. At the heart of the village stands a large, impressive Catholic church. The villagers have watched as the smaller villages in their vicinity have suffered the ravages of schism as a result of Protestant conversion (Schakt 1986), but they themselves have not succumbed in large numbers to the attentions paid to them by missionaries of various persuasions. In fact, a comfortable ecumenicalism characterizes much of religious discourse in the village. "There is only one God," my interlocutors told me. "What does it matter where you worship him'/'' Most villagers remain within the Catholic Church, although a significant minority have become Nazarenes. But the Mopan brand of Catholicism is not quite that of Rome. Prayers and offerings of pom 'copal' incense are made to the mountains, the morning star, and the sun, as well as to the supernatural 'owners' (u yumil) of the various resources of the forest and to the Christian God. The Mopan, however, would be the first to insist that their religious behavior is puuro Katoolika. An important feature of Mopan Catholicism, as elsewhere in Latin America, is the institution of compadrazgo. At the Catholic ceremonies of baptism, confirmation, and marriage, godparents are chosen for the initiate. Under compadrazgo, the godparents are recognized as entering also into an important relationship of 'co-parenthood' with the initiate's biological parents (Madsen 1967). Three mayordomos are responsible for the caretaking of the church. A system of hierarchically structured voluntary organizations traditionally took responsibility for the sponsorship of the several saints' fiestas, which took place every year. This system is in decline today, although at least one such organization still operates. A pan-village cleanup day and general meeting is held every three months, at which issues of interest to the community are raised for group decision. At this meeting, an alkaalde, or chief political officer, is elected yearly. He is responsible, with the help of a deputy and several village policemen (also locally nominated), for resolving legal disputes on the basis of customary law. In community matters, such as the building of a church or the burial of the dead, as well as the periodic chopping of bush to keep the village from being grown over with weeds, all households must contribute labor or pay a tine. The contribution of this communal labor is in fact the price of entry into the community. Such entry includes the right to farm in the area and is understood as a privilege (Danziger 1996a).
The Mopan Setting
21
A village council (an institution modeled on English village traditions and introduced to the Mopan by the British colonial government in 1965) takes a leadership role, and makes many of the other decisions necessary to the running of the village. Several other village-level societies also exist. The Toledo Maya Cultural Council, an indigenous organization with international connections that aims to unite Ke'kchi and Mopan Maya to achieve political and cultural goals, has a presence in the village. Two women's craft cooperatives, a cacao producers' cooperative, and a health committee (instituted by federally employed health professionals outside the village to facilitate their contact with it) were all part of the social landscape in the late 1980s and early 1990s. The village boasts several well-stocked shops and is a destination in its own right for travelers from the more remote villages. But there is no market here. Villagers regularly travel to the non-Mopan market town on the coast to sell their produce. Since 1948, an all-weather road has connected the village to the market town. A few villagers own trucks and can make the journey (usually about one and a half hours) on their own. Most rely, however, on the biweekly run of a local market bus to take them there and back again. The village has its own primary school, and attendance is nominally compulsory to age fourteen. Today's children reportedly attend school more regularly and for a longer period than did their parents, although complaints about absenteeism and truancy still issue steadily from school personnel (cf. Crooks 1997). A small but increasing number of Mopan villagers continue to high school and beyond. These must go to live in town or find a way to commute there from the village. Although, with the penetration of a market economy, differences in economic standing among villagers are becoming more and more evident, differences in perceived social status among them are not extreme. Villagers are remarkable for their egalitarianism and for their insistence that the same economic opportunities and advantages be available to all households in the village. The question of private title to land is becoming increasingly pressing (Wilk and Chapin 1989). For the time being, however, access to the basic resource of land for cultivation remains available to all under a reservation system that was instituted by the British (Bolland 1988) and is modulated by the local understanding that certain tracts of land belong to certain village communities, for the exclusive use of members of the village in each case (cf. Toledo Maya Cultural Council and Toledo Alcaldes Association 1997). The Mopan take the view that individuals are independent of and responsible for their own social circumstances. Mopan individuals assume responsibility for their own lives and those of their families. They carefully and politely refuse to do so for others. This perspective on the autonomy of the individual allows for a great deal of personal ambition, and many individuals work hard to better the economic situation of themselves and their families. A counterforce to this tendency exists, however, in the prevailing egalitarian ethic. If one household is perceived to have access to some resource that is denied to the others, long-standing and bitter enmities may result. Families who are doing well express fear of witchcraft that may be practiced against them on behalf of jealous neighbors (cf. Foster 1952; Gregory 1975). At the least, successful families can expect to suffer p'aas ('insult, mockery, disrespect') treatment from others in the form of backbiting and resentful gossip. In extreme cases
22
RELATIVELY SPEAKING
(which are by no means rare), the tension between individual ambition and community egalitarianism leads to violence or to emigration from the village. Emigrants may move to another village or set up house alone in the bush. These isolated homesteads become nuclei for new clusters of houses, which may in time become villages in their turn. The tendency toward fission and mutual distrust quite clearly counterbalances and at times even jeopardizes the reliance on solidary labor exchange that also characterizes relationships among households in the village. One of the functions of elaborate forms of Mopan kinship recognition, as understood by Mopan themselves, is to display to observers and to the principals in the relationship that feelings remain on an even keel, and that the particular relationship being recognized remains one of mutual respect. The Notion of 'Respect' Today, the Belizean Ke'kchi and Mopan Maya live side by side with one another and with the Garifuna and East Indians of the Atlantic coast (Howard 1975). The Mopan also maintain relations with Guatemalan Mopan-speaking communities and interact routinely with the Creole-speaking inhabitants of Belize City and of central Belize. Long-term North American visitors to the area have included Catholic and Protestant missionaries, Peace Coips volunteers, a few permanent settlers, and a steady trickle of anthropologists. J. E. S. Thompson visited Mopan territory briefly in the late 1920s and published a rich, if scattered, account of his observations there (Thompson 1930). Gregory (1984a) takes Thompson's account and the memories of old people in the late 1960s as a baseline against which to assess the changes that had taken place by the 1970s. Osborn (1982) makes use of Gregory's account, as well as that of Howard (1977 [1975]). in her own summary of life among both Ke'kchi and Mopan in Belize. I refer the reader to these accounts for broad ethnographic sketches of the traditional Mopan way of life. Osborn is the first to provide more than a cursory glimpse of domestic life among the Mopan, and her work supports Gregory's call (1984b) for further exploration of the Mopan world from the point of view of its women. In traditional Mopan society, as documented by Thompson and Gregory, status and respect were granted to individuals as a function of their service in the civilreligious hierarchy of fiesta sponsorship and of community service: The pivotal concept here is "respect." . . . The idea of "respect" has an important place in Mopan cultural tradition. And though the term might be used in a variety of contexts by present-day informants in discussing traditional customs, its constantly repeated use in the context of the civil-religious hierarchy is especially notable. (Gregory 1984a, 34)
Gregory's consultants explained to him that the reason that fiesta sponsorship gave men respect was "because they are respecting the saints" (Gregory 1984a, 36). This explanation, in turn, is linked by Gregory to his reworking (1984a, 22-25) of Thompson's original characterization of the Mopan religious world as one in which
The Mopan Setting
23
animistic deities require propitiation to ensure the well-being of humans and of their animals and crops. The individual who sponsors such propitiation does so on behalf of the entire community. Gregory (1975, 78-79) finds in this action one of the principal forces for cohesion in a Mopan world otherwise characterized by divisive social forces. The Mopan word for the "respect" to which Gregory refers is tiik. The term is still "constantly repeated in a variety of contexts," although, as Gregory predicted, the annual round of observance of saints' fiestas has gone into a decline. 2 In discussing the gradual diminution of saints' day celebrations, villagers cite the increasing expense of mounting these events now that Belizean law has begun to require the sponsor to engage a national policeman for each night of festivities. This new regulation, in turn, owes its origin to the apparently increasing drunkenness and rowdiness of the fiestas. Villagers mention violence and even murder as probable occurrences on these occasions.3 Gregory (1984a, 117) locates the seeds of these difficulties in the emergence of "a new breed" of disrespectful youth in the late 1970s. This, in turn, was presaged in 1960 by "a youthful revolt against the authority of their elders" (Gregory 1984a, 111) by the young men of the village. The "revolt" took place on the occasion of the yearly selection of the alkaalde. During the 1950s, young men had been given positions of influence for the first time, as a result of their expertise in more modern domains of commerce. In 1960, the young men evidently preempted the traditional selection of an older man, experienced in matters of custom, as alkaalde and instead nominated and acclaimed one of their own for the alcaldeship. Gregory (1984a) reports that his interlocutors spoke of this moment in 1960 as the point "when the young men took over" and "left the old men behind" (111). The incident is characterized as the opening of a new era in which appeal to traditional custom and to the authority of the elders became a matter of individual preference rather than one of socially enforced necessity. In public life, that era continues to this day. The reported previous practice of the nomination of alkaaldes by the pasaados (previous holders of the office) has fallen into disuse. Instead, nominations are made by the members of the vllage council and have become increasingly linked to issues of national party politics.4 From year to year alkaaldes may be old men, praised for their familiarity with traditional law, or they may be younger, literate, modern in outlook, and praised by a different constituency for their ability to understand the modern world and relate to institutions and events outside the village. The Ethnographer's Expectations and Experiences Having read Gregory's account before arriving in Mopan territory, I was prepared to enter a world of humming commercialism and modern values, in which the ferment of change, entrepreneurship, and departure from the old ways would be among the most tangible of my experiences. In fact my experience was quite different. Great changes have no doubt taken place in this Mopan village since Thompson visited it in the 1920s. But what struck me as a first-time visitor in 1986 was the degree of coherence of the "deep structure" of the culture (Bricker 1987) with which I was
24
RELATIVELY SPEAKING
presented. To be sure, innovation was often eyed with interest, particularly any innovation that seemed likely to augment a family's cash income or to enliven the daily round of village and household affairs. But such innovations were appropriated, Mopan fashion, into a clearly Mopan world of meanings and values—to the despair of many a development project staffperson. In short, I encountered far less in the way of modern values among the Mopan than I had expected to and I was surprised to find that most of my friends and acquaintances rarely traveled out of the village. Few of my interlocutors voiced disrespect or even doubt regarding the authority and correctness of the old ways, and many cited kostuumbre 'custom, revered tradition' as the only explanation for activities of theirs about which I inquired (cf. Reina 1966). Specifically with respect to the commercial economy that has been the motor of change in the village since 1950,1 was surprised to find that many of my acquaintances still reckoned prices in terms of use-value rather than exchange value and that they expected to control only very small sums of money in their lifetimes.-'11 attribute these attitudes to the fact that many of my principal interlocutors among the Mopan were women. A formal separation is made by the Mopan between the universe of women ich naj 'inside the house' and the universe of men yok'ol kab' 'in the world'. 6 It would be entirely inaccurate to say that Mopan women are prohibited from participating in public life. In fact, certain women do so participate, and they are not censured or hindered in any way by the community. One of the two nurses resident in the village in the late 1980s and early 1990s was a Mopan woman. A growing number of the primary school teachers are Mopan, some of them women. A few women have held jobs outside their own homes—as laundresses at nearby army camps or as domestic or sometimes administrative assistants to nurses, teachers, or missionaries who have other work to do. A certain number of village-level women's associations have recently come into existence, facilitated by non-Mopan development workers. These are supported and accepted by the community at large. Women who belong to them or who work outside their own homes are widowed or separated, or have husbands or parents who look favorably upon their involvement with the outside world. It is certainly accurate, however, to say that most Mopan women do not participate in such activity, and that neither they nor their menfolk expect or exact such activity from them. Many individual men would indeed forbid such activity in their wives. Since the forces for change have, from the middle of the century, emanated from the outside world, they have generally been a part of public, not domestic, life. They have been available to Mopan women only to the extent that their own attitudes and those of their husbands and families have allowed them to come into contact with them. While obviously palpable in all spheres, the impact of change and modernization on Mopan culture to date has been far greater in the public sphere than in the domestic sphere. Consequently, it has had a greater effect on Mopan men than upon Mopan women. In the descriptive account that appears in later chapters, I provide a brief sketch of Mopan life ich naj 'in the house'. I do so not only with the intention of providing a background to the experimental study but also with the particular ethnographic goal of introducing and illuminating a set of still-extant t~ik 'respect' attitudes and behavior that affect both men and women and that are the domestic corollaries of the respect for public (yok'ol kab') religious and political activity documented by previous observers.
3
THE M E A N I N G S OF KINSHIP TERMS
The Piagetian Definition Task Psychologists, as well as anthropologists, have had a long-standing interest in terms for family relationships. While anthropologists have been fascinated by the fact that this area of meaning seemed in many cases to hold the key to social organization, and to delineate the area in which the socially or culturally imposed met the natural or self-evidently given, psychologists have concentrated on the relational and potentially self-referential logical properties of kinship terms, especially as a window onto language acquisition and child development. Kinship terms appear on the list proposed by Ferguson (1977) of semantic domains that are universally apparent in the speech of very young language learners. Children's early usage and understanding of kinship terms (at ages one to three years) has more to do with appropriate participation in and socialization to locally relevant human relationships than it does with cognitive calculation in any genealogical grid (Schieffelin 1984; Carter 1984; Dunn and Kendrick 1982; Anderson 1977; Chambers and Tavuchis 1977; Thomson and Chapman 1977; Ruke-Dravina 1976; Burling 1970). Even in middle childhood, when children's use of such terms in discourse may be semantically adequate by adult standards, their talk about the meanings of these kinds of terms is not like that of adults (Landau 1982; Vygotsky 1934 [1962]). One of the earliest standardized intelligence tests for children (Binet and Simon 1908, 49) made use of a technique that consisted of asking children to repeat and rectify short texts containing absurd or illogical elements. One of the texts states, "I have three brothers: Paul, Ernest, and myself." Jean Piaget (1928, 62-63) became 25
26
RELATIVELY SPEAKING
extremely interested in the fact that this sentence was one that French-speaking Swiss children aged nine to twelve found harder to put right—or even to recognize as "silly"—than other Binet-Simon texts. Piaget (1928,74-88) argued that understanding of the three-brothers test depended on independent cognitive mastery of the "logic of relations" and on the ability to adopt the perspective of someone other than oneself. He elaborated the investigation by devising (Piaget 1928, 98-1.13) a set of new tasks that explored French-speaking Swiss children's developing relational and perspective-taking abilities with respect to sibling terms and to recognition of the right and left hands of the self and of others. From these parallel investigations, and referring to the substantial body of his other work, Piaget (1928, 112) traces "the non-relational character of childish ideas back to the ego-centricity of thought." 1 Relational Stages in Kin Term Definition Of particular interest to later scholars has been the task that Piaget devised at this point, requiring the children to answer for him the question "What is a brother?" (or "sister"). In his analysis of children's answers to this question, Piaget identified three stages of development. At the absolute, or categorical stage, children simply offer an absolute characteristic that applies to referents of the term, stating, for example, that a brother is a boy. At the next stage (relational stage), children recognize that the term denotes a relationship but understand the relationship as a property of the particular referent. Features of this stage include the belief that only some of the several male children of a single couple can be described as brothers, and that having a brother does not necessarily make one a brother or sister oneself. The final stage (reciprocal stage) is reached when the child supplies an answer "which implies in one way or another the idea that in order to be a brother one must have a brother or sister" (Piaget 1928, 104). At this third stage, the child understands that the status denoted is not absolute but relative— and more than that, that it involves a reciprocal relationship with some other person. Exact replication of Piaget's "What is a brother?" task with English-speaking American children has reproduced his results and supports "Piaget's contention that the tests measure developmental changes in conceptual thinking" (Elkind 1962, 131). Subsequent replication of Piagetian stages in children's discussions and definitions of kinship terms in widely different languages and cultures has since taken place (Bavin 1990; Luong 1986; Carter 1984; Deutsch 1979; Greenfield and Childs 1977; Price-Williams et al. 1977; LeVine and Price-Williams 1974). No kin term definition study of which I am aware has failed in general to support the existence and ordering of the stages as originally outlined by Piaget. Semantic Complexity and Order of Stage Progression In an exact reproduction of Piaget's task, Kurt Danziger (1957), working with Englishspeaking Australian children, expanded his study to include family relationship terms other than those for siblings. He found that these other terms underwent a developmental progression like that of sibling terms, but that all terms did not go through the stages simultaneously. A child might give a categorical stage answer for one term and in the next breath give a reciprocal stage answer for another. Haviland and Clark (1974) repeat this observation with American English speakers, and offer an explanation for it.
The Meanings of Kinship Terms
27
Haviland and Clark (1974, 33) subject the kin terms of English to a semantic analysis that operates in terms of the property of relationality considered so important by Piaget. It is based on combinations of "relational components" (Bierwisch 1970) expressing parent-child links and sets aside "property features" or absolute semantic characteristics (Haviland and Clark 1974, 29). Under this analysis the English term grandmother, for example, is analyzed as containing two relational components and one nonrelational or "property" feature (Haviland and Clark 1974, 30): [X PARENT OF A] |A I'ARKNT OF Y] [FEMALE X|
In this notation, X, A, and Y all stand for human individuals. X refers to the person designated by a given term, and Y to the person to whom X is related by virtue of the term (i.e., the reciprocal). Relational components mention two individuals. Property features mention only one. Haviland and Clark (1974) assign a quotient of semantic complexity to each term of the American English kinship paradigm. "The complexity of an entry," they explain, "is based on two factors: (a) the number of relational components in an entry, and (b) whether the relational components are all the same" (36). In their system, Haviland and Clark postulate two possible relational components: the first [X PARENT OF Y] we have already seen. The second is its inverse: |X CHILD OF Y]. In fact, Haviland and Clark embed a double relationality into these components, since each component not only specifies that a parent-child relationship holds between X and Y but also specifies which party—X or Y—is senior to the other.2 Eor Haviland and Clark's analysis this is of no importance, since no terms in the American English system use seniority alone as a distinguishing criterion. We will find it important later, however, as we proceed to the Mopan data. Haviland and Clark hypothesize (1974, 35) that less complex terms according to this measure will go through the Piagetian stages before more complex ones, for a given child. Although the complexity factor does not account for all of the variance in the results of their investigation, they conclude that degree of semantic complexity is "a very important factor in determining the order in which kin terms are learned relative to each other" (Haviland and Clark 1974, 47). Deutsch (1979, 320) offers confirmation of their claim in his own results from the German kin term lexicon, as does Luong (1986, 30) working with Vietnamese. 1 Test for Psychological Reality of Alternative Semantic Models Greenfield and Childs (1977) point out that in some languages differences in the semantic analyses offered for a particular set of kinship terms will have consequences for complexity in Haviland and Clark's sense and therefore, in these cases, will entail contrasting predictions about the order of acquisition of the same terms within a given system. In these cases, data regarding the actual order of acquisition of the critical terms could be used to distinguish among the different possible analyses that might be applied to the data.4 Deutsch (1979) clarified the point that even in a system of "kinship" terminology in which genealogy was never at stake (Deutsch devised a fictional system of relationships among rats, based on features like relative length of tail and color of coat), the
28
RELATIVELY SPEAKING
relationality expressed in children's definitions could be coded, and that the Piagetian stages in definition phrasing could still be clearly recognized.5 This means that Piagetian acquisition can be used in the manner proposed by Greenfield and Childs (1977) to discriminate even among semantic models that differ in their commitment to the presence of universal genealogical content in the kinship domain. We have already seen clearly how the three approaches to semantic meaning presented in chapter 1 correspond to different positions in the debate over the relationship between language and thought, and in their assumptions about a universal domain of genealogical kinship. Combining the insights of Haviland and Clark (1974) with those of Greenfield and Childs (1977) and of Deutsch (1979), it appears possible that the Piagetian kin-term definition task might be used to discriminate among these three approaches with respect to their applicability to a particular body of data, thus contributing empirically to the debate over linguistic relativity. My goal is to make precisely such a discrimination. In doing so it will be important to note that, in the definition task, relationship terms are processed through the Piagetian stages of relationality in an order that corresponds to their relative degree of relational complexity with respect to the adult semantics of the terms, even when the data are obtained from quite young children. This fact suggests that some approximation of the adult semantics is actually present in the children consulted, but that, in the context of the definition task, younger children find it difficult to express the relationality of more complex terms. The Piagetian task presumably yields results of the type that Haviland and Clark discovered, not because school-age children simply do not understand the kinship terms of their language but because the demands of the task are onerous enough to constitute a variety of cognitive load that interferes with normal processing. The Piagetian relationality score that is given to a child's definition of a term is thus a measure of the relational complexity of the term—but the particular definition uttered by the child is certainly not a full measure of that child's understanding of the term (cf. Hirschfekl 1989).° In this task, it is precisely the limit of what is accessible to reflective thought that serves as an index of less reflective understandings. The task can therefore be characterized as one that taps nonreflective thought despite the fact that it calls precisely for reflection upon the meanings of terms, and as one that is relevant to adult as well as to child cognition, despite the fact that children alone are consulted. Where terms are relationally complex, processing of relationality is evidently sacrificed by children to the necessity for reflection imposed by the definition task. For this reason, while I will take some interest in the content of Mopan children's definitions, definition content is not my main focus of interest. The degree of abstract relationality of children's definitions, independent of content, will be my primary concern. The Question of Kinship To conduct the test that I propose, a set of terms must be identified for semantic analysis under three contrasting approaches to meaning. It is crucial that such identification not be made on grounds that rely on any account of the notional content of the terms, since the test itself is designed to distinguish among various theories about
The Meanings of Kinship Terms
29
such content. In deciding upon the particular terms to examine in a given language, then, we confront a specific version of the "mapping problem" of semantics in general (see Quine 1965): that is, the suspicion of one's own intuitions about reality, leading to a general despair about the possibility of comparison under true relativism. How, in short, can one undertake a cross-cultural study that relies on definitions of kinship terms, when disagreement about the very existence of a comparative domain of kinship lies at the core of the theoretical discussion? We are faced with the necessity of isolating a set of terms for examination in a way that is independent of their (possible) nature as kinship terms. We must discover, without reference to any notional content, a group of candidate terms for Mopan that we can justify as a set and that we have reason to suppose will replicate Piaget's developmental results on the definition task. Linguistic Aspects of the Piagetian Task Unlike certain other kinds of nouns and noun phrases (such as natural kind terms, e.g., tiger), the lexical items traditionally known as kinship terms signify the relationship between a referent and a reciprocal, rather than simply specifying the referent itself. This is the linguistic property of "two-placed predication" (Lyons 1977,153; Silverstein 1987, 144; Bierwisch 1970). It is this semantic property that underlies the logical relationality of kinship terms in which Piaget and others have been so interested. In many cases, the formal reflex of this semantic property is not to be found within any actual token of a term. Rather, it is to be found in the term's context of use. When such terms are used in reference, two-placed predication is encoded syntactically and is to be found in the status of kinship terms as inherently possessed forms. The pronouns of possession are themselves in turn intimately linked to the speech situation, and reference must be sought in the context of utterance (Silverstein, 1976b; Jakobson 1957). When two-placed predicate terms are used in address, a particular configuration of pragmatic rules applies such that the speaker canonically stands at the position of the reciprocal.7 In combination with the possessive pronoun, then, or in address, kinship terms are like "duplex terms" or shifters (Danziger 1998; Casson 1981). As such they perform an indexical as well as a referential function in speech, indicating an object in the world that can be correctly identified only with reference to some aspect of the speech situation in which the utterance is made. The Metalinguistic Nature of the Task To score at the highest level on Piaget's task, children must deal with both the semantic (two-placed predication) and the pragmatic (identity of reciprocal encoded in speech situation) aspects of these shifter-like terms. A simple statement that a term specifies a kind of relationship scores at Piaget's relational stage. To score higher, the child must also provide an account of the term's context-dependent aspects—often including an abstraction of the self in the canonical role of speaker/reciprocal (Benveniste 1966 [1958]). The first demand, and the one that children all over the world find easiest to fulfill, is a metasemantic one. The second is metapragmatic (Silverstein
30
RELATIVELY SPEAKING
1993) and is much more difficult. Definitions of kinship terms are not always phrased reciprocally even in the academic literature on the subject. To motivate comparability of the data of the present study with other data on the cross-linguistic acquisition of kinship terms, it will not be necessary therefore to adopt a position a priori on the existence or otherwise of kinship itself, or on the semantic content of the terms involved. It will be sufficient to establish that the terms I intend to examine possess the two specific linguistic properties of two-placed predication and of shifter-likeness that underlie the Piagetian task. Defining the Domain Methods for isolating lexical sets for semantic analysis were explored and discussed at length in traditional ethnoscience, specifically to allow the analysis of terminological systems "in a way which reveals the conceptual principles that generate them'" (Frake 1969, 29)—that is, starting from terms in use and not from a previously assumed metalanguage. Such methods consisted fundamentally of establishing a linguistic environment, or "frame," into which certain linguistic units would fit, and that would exclude other linguistic units. Explicitly in the spirit of inductive discovery, the early ethnoscientif ic linguistic frame defined its own domain of interest rather than seeking to describe the linguistic arrangement of terms related by nonlinguistic or notional criteria of intuitive meaning. Ethnoscientific application of this method often consisted simply of the establishment of some single, perhaps uniquely occurring or interview-imposed linguistic environment, and what was discovered was therefore as much pragmatically as semantically or syntactically controlled (Frake 1980).8 In a more naturalistic application of a similar logic, Silverstein (1976a, 1987) uses spontaneously occurring intralinguistic contexts (such as the possibility or nonpossibility that a given form can occur with the ergative marker) to establish cross-linguistically comparable domains. My own approach to the isolation of a set of terms for consideration in the Piagetian task is similar. To identify a set of terms for semantic analysis, I distinguish, for Mopan, two linguistic frames—one syntactic and the other pragmatic—at the intersection of which a certain set of terms obligatorily occurs, and from which certain other terms are systematically excluded. Only terms meeting the requirements of both frames are considered in what follows. The Tzik Relationship Terms of Mopan Mopan Attitudes toward Reference and Respect Until very recently Mopan society has been dependent on oral transmission of its history and traditional law. Many Mopan display a strong, even religious, respect for traditional procedure. It is important to Mopan individuals that ritual and legal precedents be followed exactly, and that history not be confused with fiction, even to make a moral or instructional point (see Danziger 1996a). Fantasy and parable are little appreciated in verbal art, while a great deal of admiration and interest are reserved for the faithful rendering of traditional accounts, which are accepted as
The Meanings of Kinship Terms
31
historically true. The Mopan expression kux-kin-t-ik-0 'cause-something-to-comeinto-existence' corresponds to our create, invent, imagine, or make something up. Applied to linguistic texts, it carries strong negative connotations, equivalent to that of English lie.9 Those Mopan contexts in which a statement or object is accepted as not what it seems to be are often sacred ones. Like many of their neighbors, traditional Mopan believe that the plaster and ceramic statues of saints that adorn their churches possess religious power (cf. Reina 1966). Their theatrical dances, in which people adopt the personae of deer, jaguar, or monkey, are cheerful but distinctly sacred occasions. It is precisely the attribution of literal essence to such unlikely vessels that gives these instances some of their miraculous quality. It is considered important that certain events should be conducted as they have always been conducted in previous generations. In speeches and incantations performed at moments of religious importance—and even in everyday life when making reference to kostuumbre—the Mopan invoke the habits of "those who died before us," using terms for parents and grandparents. It is understood that, by virtue of their experience and their contact with those who have now died, old people represent an invaluable repository of information about traditional precedent. Young adult Mopan to whom I spoke repose comfortably in this belief. They are confident that the old people know important secrets of which they themselves are ignorant and that that knowledge will become available to them (the young) at the moment they need it. This confidence in the role of old people as the repositories of vital cultural knowledge is the basis for a thoroughgoing Mopan sensitivity to relative seniority in their relations with one another. Seniority is cited as a very important factor in determining the tzik 'respect' that any one individual owes to another. As Gregory (1984a) noted, 'respect' is "pivotal" in religious ceremonies honoring the saints. And the Mopan say that it is the correct deployment of tzik that distinguishes them from the animals and from other ethnic groups. On different occasions I have been told that it is tzik to refrain from sexual relations with one's spouse on the eve of planting or hunting. It is tzik to greet one's neighbor politely and to abhor anger, violence, and murder. It is tzik to listen acceptingly to one's spouse's words and to accede to his or her wishes. It is tzik to marry outside the group into which one was born, and it is tzik to honor those who have sponsored the children of others in the ceremonies of the church. It is tzik to speak the truth, to treat one's compadres and comadres with careful appreciation, and to recognize with deference the old people who know the Mopan law. It is tzik to exercise self-discipline and not to do "whatever one wants." It is tzik to bear the pain of the world with patience and to cherish and forgive one's fellow humans in it. Above all, most consistently and utterly regularly, it is tzik to greet those others who are considered to be in certain essential relationships with oneself with the term that denotes that relationship. And in strict complementarity, it is tzik to refrain from sexual intercourse with those whom one greets in this way. T:ik 'Respect' and the Greeting of Kin As a verb, Mopan tzik 'respect' is morphologically transitive. It can be translated as 'to honor' or 'to offer respect to', and it has overtly religious connotations. To kill
32
RELATIVELY SPEAKING
another human being, for example, shows acute disregard for tzik, and such disregard has direct supernatural consequences. A spate of flooding that afflicted the village in one year, for example, was blamed on the ill humor of the Chaaks, or gods of thunder and rain. This ill humor in turn was itself blamed on the perceived high rate of homicide in the village the year before. One old and learned Mopan man, when asked to describe the end of the world, told me that that calamity would be marked by the complete disintegration of tzik. Not only would men murder one another freely and take their own sisters, daughters, and grandmothers to wife, but on that day household furnishings would be transformed into wild beasts and wander away. Tzik appears in this description as a force that holds together the very foundations of the human and cultural universe. But tzik is also an oft-recurring fact of daily life and is said to be manifested in everyday greeting practice. The younger of any two persons meeting one another for the first time in a day initiates a short formal salutation, which the elder returns (see Lungstrum 1987; flanks 1990, 105). The form of the greeting is simple. Minimally, it consists of the particle 'Dioos.' Frequently, however, the greeting is elaborated by following this form with a term of personal address. Proper names are sometimes used in this position, as are terms of social office (e.g., alkaalde 'mayor' or paade 'priest'). Popular among many who consider the use of proper names disrespectful is the formulation nooch inik 'old/ great man', nooch ctiup 'old/great woman'. 1 " There are certain dyadic combinations of particular individuals in which it is mandatory for the younger person to greet the elder with the senior-party term of a two-placed predicate pair (a relationship term). The senior party then returns the greeting but without using the junior-party term in return. To put it another way, where certain reciprocal relationships are understood to apply between two people, the greeting makes mandatory reference to them. The relationships in question include those that we would render in English, for example, as parent to child, brother to sister, father-in-law to daughter-in-law, and godparent to godchila1. In some cases, greeting with a relationship term may be elaborated to encompass several lengthy turns, involving formulaic poetic speech structured into rhyming couplets. To greet someone in this way is to tzik, or to 'respect', them. Failure to carry out the minimal greeting procedure is a serious breach of good manners. Failure to address a person who is entitled to it with the appropriate relationship term is especially rude, prompting charges that one is behaving "like a dog" or "like an animal." The failure of two people of opposite sexes to greet with tzik is understood to indicate that sexual relations have occurred between them.Where a relationship term would have been expected in the greeting, this understanding of sexual intimacy is accompanied by expressions of abhorrence and anger. There is widespread understanding among the Mopan that other peoples— actual or mythological—might have the same kinds of relationships, while failing to honor them with tzik address. The fact that Mopan use relationship terms in tzik greeting address indeed, was offered to me by both Mopan and Ke'kchi consultants as one of the hallmarks of Mopan identity. The tolerance of the Mopan for alternative worldviews is exemplified in the fact that their non-Mopan comadres and compadres are indulgently exempted from tzik greeting. On the other hand, the bizarre sexual
The Meanings of Kinship Terms
33
habits with which other peoples are credited by the Mopan are considered only what is to be expected from people without tzik. Both examples illustrate the fact that tzik behavior is considered characteristic only of Mopan and not of other peoples. Two people who greet one another with a relationship term should not marry." Although lexical items that correspond to English husband and wife are in frequent use in Mopan, these terms are never used in tzik greeting. Reaction to the suggestion of such usage ranges from summary dismissal to great mirth. Inversely, married couples, or couples who have had sexual relations with one another may not greet with relationship terms. This fact of social life constrains the negotiation of marriage and ritual sponsorship ties, since these, once established, entail obligatory greeting with tzik relationship terms among the family members of the principals. My Mopan interlocutors habitually and readily talked about the reciprocal relationships that must be made explicit in the greeting as tzik. As my stay in the village lengthened, I began to tell people that I was interested in learning about tzik. My interest was understood and appreciated as an interest in the reciprocal relationships represented by two-placed predicate terms used with Dioos in daily greeting. The Set of Terms The pursuit of the Mopan study requires a set of two-placed predicate shifters, susceptible to alternative semantic analyses and coherently bounded as a domain. The terms must be two-placed predicates, to permit observation of the Piagetian categoricalrelational distinction in children. They must also be habitually embedded as shifters into the speech situation context, to permit observation of Piagetian decentration. These properties, however, are all that are necessary to ensure the comparative integrity of the particular study I propose. While I am interested in examining a set of terms that offer translational equivalents of English kinship terms, the question of denotation, or of kinship content, cannot—for fear of circularity—be a criterion for inclusion or exclusion of any given term. I begin to delineate a set of such terms in Mopan by isolating the set of those noun phrases that can be used in address, preceded by Dioosl Referring to the demand of the Piagetian task for relational terms, I take the intersection of this set with the set of Mopan common nouns that are obligatorily possessed when used referentially as the domain of study. This intersection excludes from consideration terms that are used in tzik greeting but are not obligatorily possessed in reference (proper names, social office terms such as alkaalde 'mayor' or paade 'priest' and the polite formulations nooch inik 'sir' nooch ch'up 'ma'am'). This procedure yields the set that I will call that of Mopan tzik relationship terms. Within the set are found the terms that an English speaker would use if called upon to translate almost all of those relationships that we call kin relationships, with the terms for English husband and wife conspicuously absent. The set also includes terms for certain relationships that might be considered only marginally kin by English speakers—the relationships of compadrazgo. Table 3.1 lists the terms of the set thus identified and provides, using the traditional terminology of kinship semantics (Kroeber 1909), a sketch of the genealogical range of each term, along with a short selection of possible English translations for each term. 12
Table 3.1 The set of Mopan t:ik relationship terms Term
Genealogical Range
Some English Glosses
Relationships Held from Birth (Consanguines)
tat mejen
na'
ill
ti/uid'
mam
lid' chilli
(iiiicli
xnkii' mi
kiik
it-' a H
Alter is parent/child to Ego Senior party is male Alter is senior to Ego Alter is parent/child to Ego Senior party is male Alter is junior to Ego Alter is parent/child to Ego Senior party is female Alter is senior to Ego Alter is parent/child to Ego Senior party is female Alter is junior to Ego Alter not parent/child to Ego Alter different generation from Senior party is male Alter is senior to Ego Alter not parent/child to Ego Alter different generation from Senior party is male Alter is junior to Ego Alter not parent/child to Ego Alter different generation from Senior party is female Alter is senior to Ego Alter not parent/child to Ego Alter different generation from Senior party is female Alter is junior to Ego Alter not parent/child to Ego Alter same generation as Ego Senior party is male Alter is senior to Ego Alter not parent/child to Ego Alter same generation as Ego Senior party is female Alter is senior to Ego Alter not parent/child to Ego Alter same generation as Ego Alter is junior to Ego
father
of u man: son, daughter
mother
of u woman: son, daughter
Ego
Ego
Ego
Ego
grandfather, uncle, great-uncle
of u mini: grandson, granddaughter, nephew, niece, great-nephew, great-niece grandmother, aunt, great-aunt
of a woman: grandson, granddaughter, nephew, niece, great-nephew. great-niece older than Ego: brother, cousin, uncle, nephew
older thiin I'.go: sister, cousin, aunt, niece
younger than Ego: brother, sister, cousin. uncle, aunt, nephew, niece
Acquired Relationships (Af fines)
air j'i ' .V
nn' \oo.\
iiyijiiutlo kompuaili'
komaade
Xlllll
Alter Alter Alter Alter Alter Alter Alter Alter Alter Alter Alter Alter
linked by marriage to Ego same generation as Ego is same sex as Ego and Ego are male linked by marriage to Ego same generation as Ego is same sex as Ego and Ego are female linked by marriage to Ego same generation as Ego is opposite sex from Ego linked by sponsorship to Ego
Alter Alter Alter Alter Alter Alter Alter Alter Alter Alter Alter Alter Alter Alter Alter Alter Alter
is in sponsoring relation to Ego is senior to Ego is male linked by sponsorship to Ego is in sponsoring relation to Ego is senior to Ego is female linked by sponsorship to Ego is in sponsoring relation to Ego is junior to Ego linked by sponsorship to Ego not in sponsoring relation to Ego is male linked by sponsorship to Ego not in sponsoring relation to Ego is female linked by xintl ceremony to Ego
of a man: brother-in-law
of a woman: sister-in-law
of a man: sister-in-law; of a woman: brother-in-law godfather, godparent's father
godmother, godparent's mother
godchild, child's godchild
'compadre'
'comadre' (no English gloss)
The set of terms can be characterized in traditional terms as constituting a "Hawaiian" system of kinship appellation (Murdock 1965 [1949]), with generational stratification a primary contrast in the system. A unique set of terms is in use among direct lineals (counted through sponsorship of Ego by Alter at ritual events, as well as through biological parenthood). For most others, it is appropriate and correct to use one subset of tzik relationship terms for cases in which the two parties are deemed to be of the same generation (seniority within the pair is usually also distinguished), while a second subset of terms is applied to cases in which the parties are of different generations. The same generational logic is used to distinguish among all tzik relations, whether they belong to the terminological subset of those acquired through marriage, those acquired through ritual sponsorship, or those present from birth. A certain amount of individual difference is discernible in informants' usage and recognition of terms. In no case, however, does such variation contradict this principle of generational stratification. 13
36
RELATIVELY SPEAKING
But where is the "generation" line to be drawn? Mopan girls marry and begin to have children in their early teens, and families of twelve, fifteen, or eighteen children are not uncommon. An eldest daughter is often closer in age to her mother than to her own youngest siblings. Her own children are likely to be older than many of her youngest siblings. Taking the perspective of the junior party in this kind of relationship, I asked one older woman directly what a person calls the younger siblings (itz'iin) of his or her mother. "Suku'un," she replied, or Waj yatib'-0 u-k'in, a-tataa'. Q plenly-36 3A-day lA-tatan' II he is very old, he's your tutau .
In Mopan practice indeed, individuals are often considered to be of the same generation for fzik greeting purposes if they are more or less of the same age. A woman's young children, for example, are likely to be considered of the same terminological generation as her own youngest siblings. Meanwhile, those of this woman's siblings who are closest to herself in age (as well as those older than herself) are considered to be senior generation to her children. Terminological class assignment therefore does not correspond with exactitude to the idea of generation we might form if we were to adhere to the calculation of biological relationships through parentchild links. But the difficulties do not end there. Even absolute difference of age in years does not consistently predict assignment of a relationship to a terminological class. When asked to explain, for example, why A counted as junior generation to C, while B—who was the same age as A—counted as same generation, my Mopan interlocutors would sometimes offer explanations based on criteria that indicated relative sociological rather than chronological maturity. Individual A might, for example, be attending school longer than most, or he or she might be age-anomalously unmarried. These sociological criteria were not, however, consistently applied, and could not be used as predictors of term class assignment. The observation that in Mopan Maya, considerations of experience (related, but not bound to chronological age) serve to decide whether an individual will be greeted as 'same-generation' or as 'different-generation' can be placed in the same category as those that have problematized the notion of a universal kinship domain for many anthropologists. The psychological significance of this departure from strictly biological and chronological calculation in the determination of Mopan generation will form the basis for my focused study. A purely naturalistic (i.e., genealogical) account of usage in this case fails to make sense of the cultural data. Faced with this fact, the analyst has several choices. He or she can choose to abandon the naturalistic account and its attendant universality, seeking instead a cultural account (Schneider 1968) that will supply local coherence to the ethnographic observations, at the expense of cross-system comparability. A second option is to modify the initial account by augmenting the concept of nature that is invoked, appealing to salient or central aspects of human experience. Finally, the analyst may abandon the goal of interpreting the data and contents himor herself with describing it.
The Meanings of Kinship Terms
37
For the purpose of administering the Piagetian task, a semantic analysis that corresponds to each of these three strategies will be presented in the following chapters. The first analysis is made in terms of monosemy and is oriented to discovering specifically Mopan principles of semantic contrast that will define the full range of reference of a given tzik greeting term. The second analysis is one that assumes polysemy and identifies, from Mopan speakers' reflective judgments, certain central referents of Mopan tzik relationship terms. The third analysis I have called an analysis in terms of homonymy. This analysis sets out simply to list the possible denotations of the various Mopan terms in a universalizing metalanguage that appeals to genealogical parent-child links. In making the monosemous analysis, it will be necessary to appeal to criteria that are specific to the Mopan case. In what follows, an ethnographic survey of the nature of Mopan domestic relationships provides a culturally specific translation of the Mopan notion that I have until now inadequately characterized as that of generation. In addition, in view of the fact that the Piagetian task relies upon developmental data, in the next chapters I will pay some attention to the life of Mopan children and to the manner in which tzik and the relationship terms associated with it are presented to them.
4
7Z/K AND K I N S H I P
The Native Speaker's View A set of Mopan terms has now been delineated for study, isolated from other forms of the language by their properties of semantic and pragmatic occurrence. We can call these the tzik relationship terms of Mopan and justify their membership in the set precisely in terms of their occurrence under the defining circumstances. I have considered a term as a candidate for belonging to the set of tzik relationship terms if it is a two-placed predicate that occurs in the tzik greeting frame. This is a useful and defensible procedure, and we use it to enter the domain of tzik appellation without presupposing the content of the domain. But it does not quite correspond to the Mopan view of things. Discussing these matters with a woman friend on one occasion, I made an effort to verify that the prohibition on sexual relations extended to all the individuals whom she specified as requiring greeting with a relationship term. I asked, for example, whether one could marry the children of one's godmother (who are greeted with a 'sibling' relationship term). My friend replied that one could not. "Why is that?" I asked. She replied, quite simply: Ma' jetl-ek-0 por ke yan-0 ti-tzik. NEC, be-able-SUBJ-3B because exist-3B 1 A_PL-respect It isn't possible, because we have respect (t:ik).
At this, I ventured the following direct question: 38
Tzik and Kinship
39
K'n-0 u-nu'kul a tzik? what-3B 3A-mcaning DETt:ik What is the meaning o f / reason for the respect (t:ik)l My friend hesitated. (I saw that this was a difficult question for her), but she offered this reply: Por ke ti-wefok-0. because lA_PL-rt'«A'-3B Because they are our relations (ft'ok).
The explanatory term that this woman used (ef ok} signifies a relationship of general comradeship, companionship, or alliance, which can include biological relationship. On other occasions, making similar explanations, consultants might use the Spanish-derived familia, or English family. In less frequent use, but also observed, were the archaic terms dknal and o'nel, which were also glossed for me as 'family.' 1 Conversely, the fact of habitual tzik greeting may be invoked by Mopan speakers to justify the supposition of biological relationship between two people, where the exact family connection is not known. The same supposition is made where two people have the same surname. (Persons with the same surname should greet one another with tzik relationship terms and should not marry.) The connection in Mopan discourse between the fact of being in a certain kind of essential relationship with someone and the fact of habitually addressing that person with tzik greeting is a close one. The two frames of reference are in fact readily and routinely interchangeable in talk. I might ask a question phrased in terms of essential relationships and receive an answer phrased explicitly in terms of tzik greeting. This was true to such an extent that an effort of observation and analysis was required on my part to isolate the few cases that demonstrated the separability of the two. As I came to understand matters, although reciprocated tzik address using relationship terms is always taken to indicate that the two individuals are in the relationship specified, it is not always the case that persons in such relationships actually greet one another with tzik relationship terms, even though it is ideally mandatory for them to do so. In other words, although greeting with a tzik relationship term implies being in the relationship specified by the term, the reverse is not always the case. If tzik greeting does not take place where it is expected, this is of great significance. It indicates either incest or extreme anger between the two parties. Such cases do occur, however, and the relationship may continue to be discussed in the breach. Thus, the gossip about a young woman who is said to have slept witli her suku'iin mentions the relationship—never again to be honored in tzik address—by name. The kernel of scandal in the case is that he is her et'ok, and remains so, even though these two disgraced individuals must now pass one another in the street without exchanging any greeting. Jah'ix b'a'alche'-0\ as_if animals-SB Just like animals!
40
RELATIVELY SPEAKING
This comment as originally made applied to the enforced lapse in respect greeting between the two. But the comment refers at least indirectly also to the sexual behavior, which now makes the abandonment of greeting a shameful necessity. The two acts, of sexual and of social disregard for izik, are in some sense one and the same. Tzik and Sexuality The fact of tzik greeting with a relationship term isolates those with whom sexual intercourse is always forbidden from those with whom it can be licensed. A general equivalence runs through Mopan understanding, of sexual activity with license, blasphemy, and absence of religious respect, and of sexual restraint with abstinence, prayer, and respect. Recall for example, how husbands and wives do not respect (tzik) one another with a relationship term in greeting. In contexts other than those of tzik greeting relationships, Mopan people differentially deploy sexual restraint and sexual activity to symbolic and religious effect (Thompson 1930). There are many ritual occasions on which husband and wife are said to refrain from sexual relations with each other, and others on which they deploy sexual relations for their supernatural efficacy. Sexual relations are traditionally avoided between husband and wife, for example, before the man goes out hunting. The game that is hunted is believed to be the property of divine 'owners' (;,/ yumil), who will allow it to be captured and eaten only if they are entreated with respect to do so. An old man, locally acknowledged as expert in matters of tradition, used the Spanish word respetar to explain why one refrains from sexual relations under these conditions: T:uj a-i'cspetcir-t-ik-0. OBLIG 2A-respecl-TR-INC-3B You must have respect for them.
The supernatural energies of the sexual act.which are sometimes constrained for respectful and religious purposes, can also be exploited, under what are felt to be less virtuous religious conditions. I was given the following account of the procedure a Mopan couple should follow if they wish to promote the growth of the beans in their field. They abstain from sexual relations for twelve days before planting. After planting, when the beans are in flower, the wife goes among them dressed in new clothes and with her long hair flowing. She prays first to the traditional guardians of the fields and crops (Scianto Witz', Saanto Jook 'the sacred mountains and the sacred valleys'; see Thompson 1930) that no harm shall come to the beans, and she offers incense to these guardians. But, explained my teacher, when this is done, "'you should ask the evil one to help you." He repeals: Ku a-k'aat-e' ti a kisin-i! CMPL2A-ask-SUBJ_3B PREP DET devil-SCOPE You should ask Ihe devil!
The husband has been wailing for his wife in the small shelter that stands in their beanfield. Here, he spreads a cloth upon the ground, and the couple make love. As a result of this coupling, the flowering beans will later provide fruit in abundance.
Tzik and Kinship Aay yan-0
walak-oo u-k'ux-at-a
uman
41
u-yal-il
aay exist-3B HAB-3B_PL 3A-fall-INC-SCOPE because 3A-heavy-NOM u-wicli! 3A-fruit Aay! Some will even fall over from the heaviness of their fruit!
Significant for my purposes is the enactment of the sexual encounter as a way of calling upon the powers antithetical to those customarily respected in religious observance (later in the same discourse my teacher indicated again that this act invoked the aid of "Lucifer"). The tzik relationship greeting, with its corollary of sexual avoidance, is thus part of a general 'respect' for the other that is profoundly religious in tone. Tzik and Compadrazgo In Mopan belief, careful verbal greeting and respectful sexual restraint are social and symbolic practices that necessarily apply to certain members of an individual's social circle because these are biological relations (familia, et'ok). In those cases, being in the named relationship does not depend on these respect practices. But in other cases, the obligations to greet with a relationship term and to refrain from sexual contact are motivated not by biological relationship but by relationships acquired through ritual and ceremony. It is to these cases, and especially to those of relationship through compadrazgo, that the notion of tzik respect is most elaborately applied, both through verbal elaboration of the greeting and through a higher degree of outrage at the notion of sexual relations. Recall that the relationships of compadrazgo are those that obtain between the Catholic godparents or sponsors and the parents of a young person on the occasions of baptism, confirmation, or marriage. Compadrazgo also obtains, among the Mopan, between the two sets of parents of a married couple. The relationships of compadrazgo are often those first associated by Mopan with the notion of tzik if it is presented out of context. When I introduced myself and my interest in tzik, Mopan consultants often tended to assume that I was interested specifically in compadrazgo. In one case, for example, when making a social visit, I encountered a group of other visitors who did not know me. I explained my interests to them in Mopan, using what had become a standard formula, by saying that I was interested in learning about f a n 'language', and about kostuumbre 'traditions'. One of the older women asked how long I would be staying and was pleased to hear that it would be a full year—then maybe I would have some chance of learning these things. Now using the word tzik, I explained further that I was also interested in learning about 'how you respect one another'. Where at this point some Mopan would understand my interest as pertaining to greeting, or to family relationships in general, many would make the leap at once to compadrazgo. On the occasion I am describing, for example, my interlocutor joked, in Mopan, "So you will look for a komaade here!" Meanwhile, the degree of outrage and incredulity that accompanies the suggestion that sexual relations could occur between individuals who greet with a relationship term varies with the particular relationship in question. Sexual relations between
42
RELATIVELY SPEAKING
father and daughter, between mother and son, and between full siblings, for example, are widely rumored to occur. One consultant averred in English and without much evidence of disgust or horror that full siblings might have sexual contact "in secret, and without letting their father know," although of course they could not marry. The same consultant was asked, a moment later, to consider a situation in which the parents of a husband and wife become widow and widower, respectively. Could the older people now take an interest in one another'.' My consultant took a moment even to understand the question, then suddenly reacted with shocked incredulity. "You mean kompaade and komaade could get married?!" He firmly characterized the possibility even of secret or extramarital sex in such a case as unthinkable. Finally, it is in cases of compadrazgo alone that the verbal form of everyday tiik respect reaches its most elaborate articulation. Between individuals in compadrazgo relation, the simple greeting may be amplified to a series of formulaic turns, in which the health of both parties is elaborately inquired after, and in which thanks are given to God (Dioos) for continued life and health, in the face of the constant and unpredictable presence of illness and death in the world. As in many other formal genres throughout the Maya region (Edmonson 1971), the poetic features of monotonic rhythm and the use of semantic couplets characterize this form of speech (called in Mopan kichpan f a n 'beautiful speech' or nukuch f a n 'great speech'). The skilled practice of such speech is highly appreciated in Mopan society. In many ways, then, the relationships of compadrazgo are treated as the most respected of tzik 'respect' relations. Here the mutually constructive relationship between the performance of the everyday act of tzik greeting and the existence of the kinds of relationships that call for this greeting is thrown into high relief. In some sense, it is the very fact that a high degree of tzik—including sexual restraint—is called for and maintained in the absence of biological motivation that constitutes the essence of these relationships. The Ideology of Tzik Greeting Terminology Use of the Mopan relationship greeting is usually explained by consultants as a signal, delivered in order that a genealogical connection calling for respect in the form of sexual restraint will be recognized. But the greeting, and the avoidance of sexual intimacy that accompanies it, is itself quite clearly also the offering of such respect. The greeting at once identifies a relationship that calls for respect and supplies the respect called for. Mopan individuals are extremely concerned to greet others with the right relationship term. Most such relationships are a matter of general knowledge, but, when necessary, a careful reckoning based on biological, affinal, or ritual sponsorship relations among persons, in combination with issues of sociological maturity, is used to determine an individual's being in a certain tzik greeting relationship with another. A great deal of importance is attached to this kind of reckoning, precisely because of its relationship to the social necessity for tzik behavior. Adopted children, for example, are always taught to greet their birth relatives with the appropriate term, lest they should unwittingly fail to offer respect to those who are entitled to it.2 Conversely, there is extreme unwillingness to use tzik greeting—especially that appropriate for
Tzik and Kinship
43
compadrazgo relations—under incorrect conditions. My consultants were reluctant to demonstrate the use of tzik greeting and ritual speech exchanges with partners who were not considered to be related to themselves in the manner to be demonstrated. This reluctance to violate tzik convention was presented to me, among other things, as a serious hazard to extramarital adventures. A husband might find out that his wife was having an affair in just this way, since a woman would be unable to respect her lover in greeting, if she and her husband happened to meet him in the roadway. The punishment for adultery in a Mopan woman is likely to be severe. The perceived inability of a Mopan to violate tzik conventions even under such serious threat is an indication of the more-than-practical importance that the correct application of tzik behavior assumes, even in everyday life. The correctness of tzik application is decided, in native speaker exegesis, through assessment of various factors of gender, seniority, ritual sponsorship status, and sameness or differentness of maturity. Although in usage or reflection individuals may disagree strongly over the particular term to be applied in a given case, they are unanimous in accepting the necessity for the ultimate adoption of one term or the other. The one inadmissible solution, for the Mopan, is that the situation remain a fuzzy one.3 A single relationship status must be decided upon for each dyad, and the particular terms used in greeting are understood to denote the true and essential relationship that obtains between the two individuals concerned. It is noteworthy that the final decision is based on an assessment of the relationship between the two parties involved and may or may not conform to the dictates of strict "transitive" calculation from the particularfc/fc-greetingrelationships that either person might hold with the relatives of the other (see Danziger 1996a). Because of the religious penumbra surrounding the notion of tzik, if a certain relationship term is expected in greeting, use of any other term or of no term at all would be considered not merely falsehood but something closer to blasphemy. If in certain cases, then, relationship terms are used in tzik greeting by virtue of nongenealogical criteria such as absolute age or school attendance, we must take this seriously as part of the meaning of the term in question, and more generally as part of the concept of tzik 'respect' as this is constituted in the domestic world. Mopan Life Ich Naj ('in the House') Mopan men and women cannot live easily without one another. Where a family is left completely without members of one sex or another, it is not unknown for men to grind corn or for women to farm, but it is the fact that each typically performs genderspecific tasks crucial to food procurement and preparation that forms the basis for the Mopan view of relations between the sexes. Marriage is seen less as a matter of romance, or even of sexuality, than as a matter of mutual feeding and interdependent labor. 4 The universally recognized interdependence of men and women probably underlies the notable absence, among both women and men, of any belief in the inferior capacity or ability of females, as well as of any tendency to trivialize or downgrade female work or the products of female labor. Men and women in couples make daily
44
RELATIVELY SPEAKING
decisions together, especially as they get older. Ritual and traditional offices are in fact held by couples, not by individual men.5 Women are recognized as full members of the community, on a par with men, and having as much right to the resources of the community as men. I was told (by men as well as women) that, yes, a woman could become alkaalde; women could attend the village meetings and discuss and vote alongside the men. On one occasion, in fact, I saw a Mopan woman—one of the high school-educated teachers—do exactly that, without comment or question from the assembled men. By contrast, in actual fact, most Mopan women's activities are almost exclusively restricted to the domestic sphere, and their movements and opportunities are quite explicitly controlled by family members. Although the ethic of egalitarianism and of individual autonomy applies to women as well as to men, then, and although some women enjoy great freedom of movement and variety of activity, the majority of women's behavior is controlled by others. Further, the egalitarianism that marks everyday relationships among individuals of different households in the public sphere is not visible in the domestic sphere. There, the traditional rule that made juniors answerable to seniors continues to apply. The Authority of Age Gregory (1984a) tells us that the "coup" of 1960 broke the political authority of the old men over the young. No such coup, to date, has interfered with the authority of old women over young ones, in their particular sphere. A new young bride is not considered to be a grown woman, and she is not expected to be responsible for her own household. On the contrary, it is expected that she will be untutored in many of the domestic arts, which her mother-in-law must undertake to teach her. (These will include infant care, when children are born.) The young husband continues his previous life of farm labor alongside his father, or of wage labor away from the village. The young wife adjusts to life and labor in a new setting under a new mistress. Relations between mother-in-law and daughter-in-law are extremely close, although not always amicable. The younger woman works in the older woman's kitchen, and the young couple sleep together in the house of the husband's boyhood. The girl obeys her mother-in-law as she had obeyed her mother, in particular in the matter of public outings. Imperious demands for overtly displayed diligence and industry are made, and they are backed up by the sharp tongue of the older woman. Jokes and stories abound of severe matrons who whip their daughters-in-law to keep them inside the house. In serious conversation, I was told more than once that it is the mother-in-law who originally commands her son to beat his wife as punishment for sins of laziness and waste committed in the older woman's kitchen and unwitnessed by the son. Women recount their resentment, in early marriage, of a husband's obedience to his mother, at their expense. The same women, however, now old themselves, speak fondly of the loyalty of their own adult sons, who listen to their mothers rather than to their wives. This period is obviously a very difficult one for young girls; talk of young wives attempting suicide is not uncommon. But older women recall and contemplate it with surprising equanimity. Marriage is difficult, yes, they say, but so is the rest of life. Elders seem to regard a bride's tears at marriage with something of the attitude that
Tzik and Kinship
45
1 might take toward a child who cries on the first day of school. Distressing, yes, and regrettable, of course, but the procedure is inevitable, and everyone knows she will get used to it (suktal). In particular, members of the family of origin of a young bride are unwilling to speculate publicly about the possibility that she may be in distress. The Authority of Gender In the only social relationship in which gender is specifically linked to power, wives are answerable to husbands. According to verbal report, physical violence of husbands to wives is so common as to constitute something of a norm. Although this is a focus of concern in the community, both for men and for women, it apparently is not a recent phenomenon. Even some of the oldest of my informants report having suffered extreme violence at the hands of their husbands in their younger days. One elderly widow expressed her surprise at having outlived her husband. She had always been sure he would eventually kill her. Women have certain defenses open to them against brutality from their husbands. Under traditional Mopan law, a husband can be fined if he beats his wife too severely. Older women may simply stand up to their husbands, leaving to set up house on their own account or to live with a grown child. Some women, I was told, turn around and beat their husbands right back.6 In Mopan life as I observed it, however, in general a husband is free to discipline his wife with physical force. Beatings are apparently nominally administered as a punishment for idleness or for any perceived forwardness on the part of the woman that a husband takes for incipient infidelity. For some women, the mere fact of being seen out alone on the public road can be enough to precipitate violent retaliation from the husband. The fact that these attacks usually occur when the men are drunk renders them no less terrifying to the women involved.7 If a young bride is too unhappy with her situation, she may take flight from her husband's family and return to her own parents, perhaps taking her baby with her. This usually happens early in the marriage; it is seen to be a realistic option only as long as there are not too many children. It is a drastic step for a young wife to take and usually produces contrition and promises to change on the part of the husband— backed up by his family, who have an investment to protect in the marriage (see later discussion). The girl's parents, by contrast, have an ambiguous role to play here. Her mother is likely to be delighted to have her daughter back, but her duty to the other family (and, ultimately, her continued hold on her own daughters-in-law) requires that she cooperate in returning the girl to her husband and his family. In some cases, however, the girl does not return. She remains at home with her children or goes on to establish a new relationship with another man. This is an outcome about which young men express great anxiety, especially in view of the financial investment that they have made. Some say that it would be better to live with a girl first for a year, to see if she were going to run away, before getting married. Because of the danger that the girl will abandon the marriage at its outset, the mother of the bride adopts a posture of elaborately displayed consideration for the mother-in-law's claim to the young girl and to her labor. Exaggerated exchanges of respectful speech demonstrate the continued good relations between the two sides of
46
RELATIVELY
SPEAKING
the family (now komaades and kompaades to one another). A mother is cautious about visiting her daughter and about encouraging visits from her (which in any case should be conducted only with the mother-in-law's permission and company). When these visits are made, the mother is careful to avoid giving the impression that she and her daughter have any private concerns whatever, to the point of ostentatiously avoiding being alone with the girl. Admonitions against too frequent or too lengthy visits to one's mother are included in the formal advice (tzeck) given to a young girl at her wedding. Work This early period of difficulties notwithstanding, most Mopan couples continue to live together and to achieve their fair share of happiness. There is a Mopan expression that is translated as English love: Yuj u-yub'-ik-0. acute 3A-feel-INC-3B She 1'eels it strongly. This expression is used to characterize the relationship between loving newlyweds, as well as for the relations between parents and children. However, Mopan tend to characterize the relation between a married couple of long standing somewhat otherwise, although no less strongly. A Mopan woman, for example, might say of a long-standing relationship with a certain man that she is 'accustomed to him'. Suk-aj-en ti'-i'. be_accustomed-INCH_COMPL-lB DAT-3B 1 have got used to him. This expression conveys a deep sense of linkage, and in particular, that this person's absence would be keenly felt as loss. It is a bond established through steady association that is not easily broken. The same expression (suk-aj-o'on ti'i 'we are used to them') is used by Mopan to refer to their attachment to their revered traditional customs and law. Sustaining and underlying the relation of mutual feeding that provides the framework for thinking about relations between the sexes is the fact of arduous and quasicontinuous physical labor on the part of both men and women. Women's labor consists of house-maintenance tasks (sweeping, laundry, the drawing of water), some wild food gathering, and the minding of domestic animals. It consists especially, however, of food processing. The preparation of corn for tortillas is the primary task. Particularly fatiguing and monotonous is the necessity to grind by hand whatever is to be eaten that day.8 Industry is one of the principal values of the community and 'laziness' (sakan) the commonest slur. The Mopan woman demonstrates her virtue by rising early and working hard at traditional manual labor throughout the day. Even when the heavy chores are done, the Mopan woman will be observed sitting at embroidery or the manufacture of kitchen utensils rather than completely idle. Moments of rest and
Tzik and Kinship
47
respite are available only at times when she knows she is not under scrutiny from authoritative others,9 or when she is ill. Illness is an ever-present actual threat and is the focus of much concern among the Mopan. Specialists in traditional healing offer their diagnoses and services, which are widely trusted. But death is never a remote possibility, and it is mentioned freely and frequently, even in the presence of children. Illness constitutes one of the few legitimate respites from continuous labor, and extremely hard work is also understood to be a cause of illness and fatigue. Complaints and laments about one's weariness and one's symptoms of illness therefore perform the double function of calling attention to one's virtue in having worked hard enough to get sick, and of excusing one for whatever lapses in industry one might currently be exhibiting. Illness and death are accepted as frequently occurring and inevitable facts of existence, arising partly from the necessity for constant labor on the part of human beings. Tzik Relationships and Their Socialization Early Life Mopan women frequently remark on the arduousness of marriage, and of the labor involved, particularly as the family grows and before any little girls reach an age at which they are able to assist. The birth of a girl child is often announced as that of i.\ jiich' 'a girl (who) grinds'. Women say it is better to have daughters than sons. Girls will help you, whereas boys just make more work. Birth control is largely unknown, and families of ten, twelve, or more children are common. For the most part, children are desired and welcomed, although the increasing economic burden as the family grows larger is certainly a cause of anxiety. Women may find solace from the difficulties of marriage in the company of their children, particularly tiny babies. "A baby is like a toy," explained one woman to me; it can be a source of joy and comfort when everything else looks bleak. Women who have no children are pitied, and there is some fosterage on this basis—a woman with many children may give one to another who has few or none to raise. This is fairly rare, however, primarily because even those with large numbers of children are unwilling to give any away. Mothers carry babies with them wherever they go, in a cloth hammock that hangs on the woman's back, supported by a band around her forehead. When a baby cries, an effort is made to discover what is wanted and to supply it."' When not on the move, the baby is cuddled or put to sleep in the hammock or on a mat on the floor. Once at the work location, a premium is placed on the baby's ability to lie quietly and alone for fairly long periods to allow the mother to get on with her work. A baby talk register (Ferguson 1977) is used by both men and women in speaking to infants. It is characterized by high pitch, special lexemes, and much repetition. It differs from White American English baby talk in the fact that it does not provide conversational openings for the baby. Open and tag questions are not characteristic of this register; neither are pauses and interpretations of utterances. The Mopan baby is not asked what it is doing; it is told.
48
RELATIVELY SPEAKIMG
In general, the baby's utterances and actions are interpreted with amusement as versions of adult activities. The waving of a baby girl's arms is her attempt to grind corn; those of a baby boy to chop wood. A popular game with girls is a form of pattycake that imitates the pressing of tortillas onto the griddle. The vocalizations of infants are sometimes described as 'cursing' (k'ey), 'singing' (kay), or 'greeting' (tzik). Despite these characterizations, I did not notice much interest in the contents of these early vocalizations, or any consistent attempt to interpret them as words. Infants and children are also frequently characterized as tz'iik 'angry' and are playfully accused of 'lying' (tus) when their demands are perceived as inconsistent or unclear. Note that tzik greeting, like lying and cursing, is one of the adult linguistic activities with which infants are playfully credited. Tzik relationship terms are introduced to children very early in life. The terms na' 'mother' and tat 'father' (sometimes modified with the diminutive suffix -ich) commonly occur in Mopan baby talk register, as playful terms of endearment for the child. A baby girl is frequently and repeatedly addressed by adults as in na' 'my mother'; a baby boy as intat 'my father'. 11 These forms of address are normally, but not always, abandoned with baby talk register. An only child or the youngest child of a large family may bear this form of address as a nickname, used by all in the community, into adulthood. As juniors in society, Mopan children do not receive the tzik greeting. In fact, the existence of reciprocal tzik relationships is not always even attributed to very young children. One family I knew well became very merry over my suggestion that a baby under a year old could be kiik 'senior same-generation consanguine/elder sister' to another, a few months younger (the mother of the first baby was the nuclear family sister of the second baby). Although the elder siblings of the two babies were habitually recognized as related, and although it had been agreed that both babies could be itz'iin 'junior same-generation consanguine/younger sibling' to these other children, the family ultimately refused to endorse the idea of an infant as kiik even in principle. They cited the inappropriateness of giving so much respect to a very young person. Since tzik greeting is expected to be initiated by juniors, however, and since it is juniors who must use relationship terms in greeting if they are appropriate, tzik greeting and the correct production of tzik relationship terms are routinely expected from children, including the very young. Reinforcing the link between absolute age and the respect greeting, I noted that it was acceptable and even encouraged for young people to greet a wide variety of older adults with a term of the set 1 have so far isolated. In the opinion of one learned and highly respected old man, children should greet all old people with tzik terms, to show their generalized respect for age (consultants disagreed on the suitability of this practice). Again, when I asked Mopan individuals to reflect on and talk about those whom they greeted with tzik relationship terms, I often encountered the situation in which an individual stated that the junior term did not apply to a particular dyad, when the senior term had already been quite readily supplied. Conversely, children do not generally use the tzik greeting among themselves, although older children are aware that they may be entitled to it from their younger playmates. Adults say of children that they are sudak 'shy, modest, humble, ashamed' to receive the greeting, with its burden of cultural respect.
Tzik and Kinship
49
During the first year, most Mopan babies undergo two important rituals. The first is Catholic baptism, at which the baby is initiated into the church and acquires a pair of godparents. The second ritual, which takes place at the age of about six months, is similar to ceremonies described for the Yucatan (Redfield and Villa Rojas 1934). On this occasion (called the xuiil ceremony in Mopan), an adult of the same sex as the child (not the parent) shows to it the objects that it will use for its work later in life. The adult speaks to the child throughout and finishes by spreading the baby's legs astride his or her hip. In early infancy, and before this ceremony, the baby's legs are tied together, in the belief that this helps them to grow straight. It is also believed that if the baby's legs are allowed to spread before the ceremony occurs, severe diarrhea (a common killer of babies and young children) could result. This demonstration to infants of the tools of gender-appropriate labor was described to me as u sekreetajil kuxtal 'the secret of life' and is indeed treated as secret and important information. A permanent and somewhat formal relationship is established by it between the child and the adult who performs it. I have been told that this relationship entails the obligation to greet with a relationship term, and I have observed such greeting once or twice. Beyond the enactment of the original ceremony, however, the relationship is not highly visible today. 12 As the baby girl in the household in which I lived approached the age of one year, I was told that she was beginning to kaal winikil 'acquire maturity' and to gain understanding of the things around her. She now recognized individuals and activities, I was told. At about the same time, I noticed a new kind of speech addressed to her. The fond baby talk register was maintained, but at times now I would hear an order or a reprimand addressed to her, in the sharp peremptory tones that I had grown used to hearing spoken by adults to older children. At this stage, the baby was beginning to be held accountable, as older children are, for crimes of waste (spilling something, or asking for something not consumed), especially if the caregiver was likely to be held ultimately responsible for the loss (a similar family dynamic is described for the Yucatan in Gaskins and Lucy 1986). Sometime shortly after the first year, most Mopan children will acquire a younger sibling. The older child is weaned when the younger one is born and must start to walk the long distances that he or she still travels with the mother while the new baby is carried. When the child cannot accompany the mother, another caregiver is found for it, perhaps an older sibling (boy or girl, from the age of six or seven) or the father during slack agricultural periods. Child caregivers are socialized into an attitude of tender solicitude toward the itz'iin 'junior same-generation consanguine/younger sibling', which endures into later life (cf. Schieffelin 1984). Even in the very early years, Mopan children are remarkable to Euro-American eyes for their self-possession. They conduct their own affairs without much supervision from adults, and they do not rely on adults to amuse them. Children are not used to engaging in conversation with adults and often subside into silence when addressed directly. They are, however, extremely affectionate during early childhood and unabashedly approach those they know for cuddling and physical play. Infants and children alike do without elaborate toys. One of their greatest pleasures lies, as it does for children elsewhere, in the imitation of adult activities. Mopan children are allowed to play alongside adults and to participate in the tasks that adults
50
RELATIVELY SPEAKING
carry out. In the second year, children are already being praised and encouraged in their efforts to perform minor adult tasks such as shelling corn. By about age three or four, children are of genuine assistance in the house. At this age they are called upon to run errands (including shopping trips that involve responsibility for precious reserves of cash) and also to provide companionship and chaperonage to women on their outings to village and farm.1-' Middle Childhood As childhood progresses, children eagerly insist on carrying out new tasks. Learning occurs predominantly through observation and example, rarely through formal teaching. I never saw an adult push a child to attempt something new, but I was constantly being surprised by the precocity of children at performing the various tasks of daily life. Adults told me that although it is necessary to teach children how to do things (by showing them what you yourself are doing), it is not necessary to force them into labor. But children are reprimanded for other things. The principal crime at this stage remains that of waste. Little, if any, attention is paid to the intentions of the child who spills a dish of food or who loses a piece of property. The severity of the punishment matches the value of what is lost more closely than the degree of malice attributed to the child. Children do not attempt to excuse themselves on the basis of good intentions, and more often they deny responsibility or blame others to avoid punishment themselves. I have heard caregivers ask children, in rhetorical reproach at waste, "Don't you have any respect?" Ma' waj yan-0 a-tzik? NEG Q exist-3B 2A-f;;7c Don't you have any respect (t:ik)'1
As Mopan children get older, parents and other caregivers explicitly and repeatedly inform them of the term with which they must greet the various older people in their lives. In one of the rare instances of formal teaching to Mopan children, greetings are prompted or coached, with caregivers offering the correct phrase for children to repeat. These coaching sessions represent more than just practice, however; production of the utterance by the child has performative weight, and caregivers will not willingly abandon the effort until the appropriate phrase has been produced by the child. (The situation is not unlike that of a middle-class American parent coaching a child to say please or thank you to a visitor.) Usage that is considered erroneous is quickly corrected. Mopan children have abundant opportunity, therefore, to learn exactly who among the interlocutors they meet in daily life they must greet with a relationship term and precisely which term it should be. The appropriateness of a certain term to a certain specific relationship is established, for the child, through a historical chain of reference, very much in the manner that has been described for natural kind terms (Kripke 1972, 328).14 By about age six, children are expected to manage the greeting without coaching or modeling, and they are held accountable for lapses. Children do not undergo
Tzik and Kinship
51
ceremonial introduction on ritual occasions and do not use the affinal set of terms. They are often taught instead to greet the spouses of their older brothers and sisters as sukit'un 'senior same-generation consanguine/elder brother' or kiik (kikij) 'senior same-generation consanguine/elder sister'. Sometimes this is reciprocated with itz' /in, but usually it is not. Such usage may change at the child's own marriage, with ceremonial introduction to a new tzik status (see later discussion). Toddlers' tears are still silenced with indulgence but sometimes also with fantastic threats (The cat will come and scratch you if you don't stop crying! The nurse will come and give you an injection! cf. Brown in press). It becomes important for children to obey adult commands quickly, silently, and efficiently. Children who have trouble doing so are labeled 'hardheaded' (chich u pol), an expression that has negative connotations, and is also used to describe a person who cannot learn easily (once again, factors of intentionality in the child's disobedience are not considered relevant). Children who do not obey or who are loud, rude, or generally undisciplined are also scolded as tzimin 'horse, tapir'. The dominant mode of adult address to children becomes the shaip tone of command or rebuke. The threat of blows is common, although these are only rarely administered. In general, small children are interested in participating in the world of adult tasks and are proud of themselves when they are able to do so. Their tasks are learned informally, through observation and imitation. Adults allow children to watch freely and will carefully show a learner what they are doing. Schooling Almost every child in the village attends school intermittently from the age of about six or seven to that of twelve or fourteen. The community supports and nominally appreciates the school. It is a part of every alcalde's yearly round to appear there and to exhort the children to attend conscientiously. There is a growing feeling that education is in principle a way for children to achieve the prosperity that the Mopan see outsiders as having (see Crooks 1997 for more on this point). But educating one's children is expensive and troublesome. Often they are needed at home (usually for child care) during school hours. Their labor is definitely needed after school, and time and space for homework are rarely provided. The children themselves in general prefer the activities of home and farm to those of the school. Children's early training in obedience, however, and a genuine willingness to do what adults ask of them, keep them engaged and attentive to a surprising degree, even under quite trying conditions. The village school, which is one of the largest in the region, is staffed by conscientious but often frustrated young teachers (some of them Mopan) who struggle with the difference between their own values and those of the community. Classes are large, and equipment is poor and scarce. Much of the learning consists of rote memorization and the copying of set lessons from the board.15 Corporal punishment is a part of school life. It is feared by the children but is accepted as inevitable. Formal teaching and learning also occur in certain settings distinct from that of the school. Certain domains of knowledge, notably those ofpiiliyaj 'witchcraft, obeah' and ilmaj 'seeing, healing' are reportedly learned through formal apprenticeship that culminates in an initiation event. In matters of traditional ritual and law, younger adults
52
RELATIVELY SPEAKING
seek out older ones for explicit instruction (e.g., how to conduct themselves as compadres for the first time). At marriages, and sometimes also at confirmations, older people come forward and deliver format homilies (tzeek 'advice, serious business') to the young people on the correct manner in which life should be lived. In each of these cases it is understood that old people possess the requisite knowledge, and that young people are in the position of learning it. Traditional knowledge is understood, by young and old alike, to be unequally distributed on the basis of age. This confidence in the wisdom and experience embodied in absolute age is the basis for the generally stated attitude of respect (tzik) for the old, codified in tzik greeting norms. As we have seen, it is always the younger individual who initiates the greeting, and who uses a relationship term in greeting the elder. Late Childhood As Mopan children continue to grow, the tasks expected of them in the home begin to diverge on the basis of gender. At about age seven or eight, little boys are taken by their fathers to the fields to learn a man's work. Sometime later (usually by the age of about eleven or twelve) they will be given their own field to plant. Little girls remain with their mothers, but their tasks, too, expand. They now must begin to contribute their share of muscle power to the daily corn grinding, and, a little later, they must learn to shape tortillas and cook them on a hot griddle. Older children take an interest in tzik and ask explicit questions about family relationships and about the appropriate tzik term to use in a given context. Even in late adolescence and into early adulthood, young people turn to their elders to inquire about correct tzik greeting practice in unfamiliar cases. Adult talk to older children about tzik address and tzik relationships is "explanation-rich" (Linde 1987) and expansive. Adolescents have explained to me the process of an almost mystical enlightenment as they become aware of their biological connection to individuals whom they had been virtuously addressing with a tzik term ever since they could remember. What is being learned is the adult exegetic view of tzik greeting, that certain relationships are honored in greeting ultimately because they are relationships of preexisting essence. This knowledge is infused with the sense of religious importance that correct tzik address has had for these children since their infancy. It is at this stage, and as children approach adolescence, that the first charges of being 'lazy' (sakan) appear and that punishments with scoldings, threats, and sometimes beatings occur for what is characterized as unwillingness to work. Many women recall this as a harsh period in their own lives but reflect, from the vantage point of age, that their caregivers had in fact done well by them, since they had learned to work hard. Current caregivers characterize and justify their behavior in similar fashion. Children must learn to work, they say, even if they do not want to. This, I was told, is not for any reason of future proficiency but because of the imminent possibility at any moment of the parents' death. "Who will feed them if I die?" I was asked rhetorically. "Who will take them in and feed them if they cannot work?"
5
C R E A T I N G 7Z/K RELATIONSHIPS
Ritually Established Relationships My informants' immediate responses to my inquiries about tzik often had to do with the acquisition of tzik relations through ritual sponsorship or with the maintenance of such relations through the special speech genre called kichpan fan 'beautiful speech'. As we have seen, such ritually established relationships, especially the relations of compadrazgo, remain in various ways (solemnity of everyday encounters, elaboration of the tzik greeting in the form of kichpan t'an, and accentuation of the degree of abhorrence in contemplation of illicit sexual relations) the most ?z/£-like, or most highly respected, of Mopan tzik relations. That this is so highlights the tension between the ascriptive and social-interactional meaning of tzik terms in Mopan life and their alternative biological-procreative meanings that is at the heart of this book. Godparents and Godchildren In previous accounts (Farriss 1984), the institution of compadrazgo ('co-parenthood', on the basis of ritual sponsorship of children in the ceremonies of the Catholic Church) has been discussed as an emergency adaptation of Spanish practice to the alarming decline in indigenous population that followed the conquest. It came into being in its New World form, it is argued, to provide insurance for the upbringing of children in a world where adults died with devastating suddenness and frequency. The godparent remains an extremely important figure in Mopan conceptualization. It is still said 53
54
RELATIVELY SPEAKING
that the godparent is a kind of parent (see later discussion) and, as such, must make sure that the child is doing right and guide him or her in doing so. Mopan godparents have a certain amount of authority to socialize and discipline the child, as well as some right to the child's labor and a responsibility to supply certain of the child's material needs. The homilies delivered by godparents on public occasions constitute one instance of traditional teaching in the formal mode. Godparents are sometimes selected for their perceived present and future advantage to the child in terms of connections and wealth. For this reason it is quite common for Mopan families to undertake godparent relations with non-Mopan individuals. Many of the high school students from the village, for example, make use of non-Mopan godparents in town to solve their boarding problems. Alternatively, and more often, godparents are selected for their status as examples of good Mopan lives correctly lived. Ideally, they are people who are perceived as having experienced the major Mopan life events and who await "nothing else before death." The virtues of industry, piety, sobriety, and fidelity are sought in godparents, and the formal advice that they deliver to their godchildren makes explicit and repeated reference to these qualities. Godparents may be chosen from among those whom the parents already greet with a relationship term. In advance of the day set by the church for the baptism, confirmation, or marriage ceremony, a formal visit is made by the child's parents to another individual or couple to make the request that they serve as godparents. The special genre of formal speech, known in Mopan as kichpan fan 'beautiful speech' or nukuch fan 'great speech' is used on these occasions. This kind of speech is structured in parallel semantic couplets and marked by a readily recognizable and monotonous intonation. The request is framed as a request for 'work' (mevaj) that the godparent will perform tor the parent. The godparent also refers to it as such. If the request is accepted, the participants consider themselves komaades and kompaades from that point and begin to address one another as such. The child also, from this point on, has an obligation to address the godparents in respectful greeting as na'yoox 'godmother'and tata'yoox 'godfather'. It is extremely important that children give respect to those who are entitled to it with this t~ik greeting. The new godparent has undertaken minimally to pay the church's charge for the ceremony and to supply a candle and a white cloth for the child's use during it. After the ceremony, the godparent is invited to a celebration at the home of the child's parents. The lavishness varies, but the meal served will feature meat, and a certain amount of female labor from outside the household will probably have been necessary to prepare it. At this gathering, compadrazgo relations are established between the principal adults and the ascending lineals of each side (i.e., the parents and grandparents of my komaades are also my komaades and kompaades). An exchange of kichpan fan, complete with couplet structure, takes place when the godparent arrives at the house, when food is offered and accepted, and again when the godparent leaves. The parents thank the godparent for having agreed to take on the work that had been requested, for not having swept the request aside, or having 'spilled it like water'. Ma'
in-pdl-uj-0,
in-tox-aj-0.
NEC, I A-discard_solid-COMPL-3B 1 A-t1iscard_liquid-COMPL-3B I did not sweep it aside; I did not slop it away.
Creating Tzik Relationships
55
The godparent acknowledges that he or she has indeed accomplished the promised work and accepts the food that is offered.1 The responsibilities of godparent to child once the ceremony is over are somewhat vague and depend largely on the goodwill of the persons concerned. As the child gets older, he or she visits the godparent, bringing food from the parents and offering to help around the house. The godparent will try to provide a small gift in return. In my observation, relations between children and their godparents were in general characterized by deference and a certain amount of fear, but a range of relationships is of course known. In some cases godparents provide a welcome outsider's perspective on family problems and can be of great support to the child. Co-Parents As in other Latin American societies, the cultural weight of the godparent relation among the Mopan falls not on the interaction between godparent and godchild but on that between the child's parents and the godparents. Of all of the relationships that are respected by name in tzik greeting, it is this one that calls forth the most pronounced performances of tzik behavior. Relations between komaades and kompaades are taken extremely seriously. Individuals who have this kind of relationship should always speak the truth to one another and should never laugh together. Sexual contact is strictly forbidden. Enemies are easily made in this face-to-face society, but should komaades and kompaades become enemies, it is a grave matter, contrary to the spirit of respect that unites them. Formal visits mark the relations between komaades and kompaades, and a special subgenre of kichpan fan continues to be exchanged between them. This is the k'aat tojoolal, literally an 'asking as to health'. On visits between komaades and kompaades, long, repetitive speeches are ideally exchanged that felicitate the two parties on being in continued good health and that comment upon the suddenness and inevitability of sickness and death. Tzik is frequently mentioned in these exchanges as a desirable personal quality that endears one to God and also as an essential ingredient of the compadrazgo relation. A short excerpt from one recorded exchange gives the flavor. Judi' ilik a tojoolala. Ki' it yub'cib'al, ki' uxikininbil yok'lal jun tun/ fi t:iktnaj. Jah'ix a/aa ti tan ti wi/ik, ti chaantik ti b'ajili. Ma' Jan ti k'atintik waj yajil, waj k' ojaanil yok'lal jun tutil ti tzikmaji. Tojoolal ilik kitchij a ki' u yub'ab'iil, 11 xikininb' Hi. Leek ab'e' a mas t:aja, porke jun tuulik Dios, kichpana, ku t:'aj u tojil tak ti woolc'ex a wetel kompaade. "Good health is only this. It's something good lo learn, good to hear, about our compadres. Just as we are seeing, regarding one another today. We do not desire any pain, any sickness for our compadres. Only good health is good to learn of, good to hear about. That is the most important, because it is only God, the perfect, who gives health to us, compadre."
I am told that people extend the length of these kichpan fan conversations as much as possible for fear of "people saying that they are bad." Extended respectful exchange of this kind reassures not only the community at large but also the komaades
56
RELATIVELY SPEAKING
and kompaades themselves that there is neither sexual intimacy nor bad feeling between them, and that the relationship is being respected as it should be. The everyday perils of misunderstanding and anger between individuals are magnified when the individuals are in compadrazgo relation to one another. During the exchange of kichpan t'an, the participants promise not to p'aas 'insult, mock' one another, 2 and instead, to tzik each other all of their days. In particular they promise to do this by always honoring the obligation to make a formal greeting to one another. It is unequivocally good to be asked to sponsor another family's children and to become komaadelkompaade with them. People who have received and honored many such requests are proud of the achievement because it represents recognition of adherence to exemplary values and good standing in the community. To be asked to serve as a godparent is a form of prestige that members of the community can bestow on one another. As one consultant explained it (in English), "You feel good when you are a komaade. You have more respect." Acquiring Relationships through Marriage As a woman ages, she continues to bring forth children, and her eldest children take a great deal of responsibility for their younger siblings. Adolescent and teenage girls provide much help to their mothers, especially since by this stage they are beyond the demands of school attendance. Grown daughters take the weight, as well, of the everlarger burden of domestic labor. The women of the family now probably feed several adult men, as the older brothers repair to the fields with their father every day. But a daughter is not a permanent asset. As she enters her teens, the young girl is seen in public less and less, and she becomes the object of marriage (and other) proposals from young men. After she marries and leaves home, her mother will be short one hand to help. If a recent bride has no grown sisters, it will become more and more obvious to everyone that one of her grown brothers must now marry. Many women to whom I talked stated that they had married because their husband's mother was unable to manage her work alone, either because of an imbalance of males to females in the household or because the older woman had become ill. Informal Marriage As one young male consultant told me, "There are many ways to get a wife." The simplest, and the least expensive from the point of view of the boy, is to "steaF'one. This involves a boy meeting a young woman secretly, perhaps while she is washing clothes or dishes at the river, and engaging her affections over the course of several meetings to the point where she is willing to run away with him. Certain young swains make use of magic spells administered by specialists to accomplish this. 3 The parents of young girls (especially fathers) are extremely concerned to guard against this possibility and may react violently toward both young people if they discover that plans of this sort are afoot. It is therefore quite a dangerous undertaking to plot such an elopement. Difficulties and pitfalls notwithstanding, these matches are frequently made and are successfully carried off in many cases.
Creating Tzik Relationships
57
The young woman then goes to live with the boy and his family without benefit of matrimony. In some cases, the couple begins to save for a formal wedding, which may take place up to several years later. In others they dispense with this ceremony altogether. Where no ceremony is carried out, there is no use ottzik relationship tenns, either in reference to the partner's family or in greeting them. Women who are not married to the men they live with may feel themselves to be at a minor social disadvantage. But the real disadvantage to this arrangement for the woman comes in case the man begins to abuse and beat her. When a woman is not legally married, there has been no expense on the part of the husband's family to obtain her and therefore no necessary interest on their part in retaining her by restraining the excesses of the young man. Quite apart from that, however, her own family of origin often refuses to allow her back into their house. "You chose him yourself," they are reported to tell her. "Now you live with him!" Formal Engagement and Marriage Traditionally, however, and still most commonly today, the parents of a young man interested in marriage visit those of the girl three times to make the request that they allow him to marry their daughter. The girl's parents are usually difficult to convince, and refusals are not at all uncommon, especially if the girl is very young or the suitor lives far away. Refusals are couched in terms of the youth and inadequacy of the girl for marriage, and in particular her inability to do household work independently. 4 The request and its acceptance or refusal are carried out by means of lengthy exchanges of kichpan f an. Some families engage an expert at this kind of speech to make the request for them. During the entire proceedings, the young people may not contact one another. I was told that the girl's father might shoot the boy if he tried to speak to her at this stage and that the girl might even be unaware that anything out of the ordinary was in the wind. The parents of a girl give careful thought to the acceptance of a suitor for their daughter and show a great deal of emotional concern when the time comes to give her up to him. In some families, a girl's opinion may be solicited regarding her acceptance of a suitor. But young Mopan girls are not schooled in resistance to authority, and refusal of a suitor favored by parents is rare. Girls quite trustingly accept their parents' choice, all the more so since most are ignorant of the facts of sexual intercourse and reproduction.5 Older women speak of the feelings of anger and betrayal that they felt in early marriage toward parents who had consigned them to a violent or drunken husband. Having let it be known that it is time for him to marry, the boy's parents, on the other hand, usually allow him to tell them which girl he would like them to ask for. As long as she is healthy and, above all, hardworking, they are likely to accede to his choice. The constraints of the ?z/fc-greeting and its relationship to sexuality must be taken into account, however, and these are not always simple. The two young people to be married must riot be related to one another before marriage in a way that requires greeting with a relationship term. In addition, the young people's parents and grandparents must be eligible to enter into compadrazgo relations with one another. A certain girl, for example, was eliminated as a marriage possibility for one young man because it was widely known that his father and her mother had en-
58
RELATIVELY SPEAKING
joyed a brief sexual liaison many years before. That earlier relationship disqualified the older people from ever entering cornpadrazgo relationship with one another, and the young man was told that he must choose a different girl for his bride. It is considered very modern for a young couple to modify the traditional request procedure slightly by engaging in a written correspondence before the formal request is made. The young man tries to discover, by letter, whether the girl is likely to look favorably upon his suit. Only if he gets some reassurance from her will he approach his parents lo undertake the formal arrangements. 1 was told that in some families a girl who receives such a letter will tell her parents about it and let them know obliquely if she favors a young man's candidacy. Parents are not averse, in principle, to marrying their daughter to a man she likes, and these marriages have become more and more acceptable. Certain adults see this practice as the wave of the future and look forward to a time when all village marriages will be made in this way. A girl's parents often refuse the first few suitors who come their way and evince great reluctance to marry their daughter off. Eventually, however, a suitor will be accepted. The young man's family press their case on the basis of the natural and universal interdependence of man and woman and on the necessity for every adult to find a partner who will join with him or her in producing food and drink from the land. B'aalo ilik walak u yantul, b'aalo ilik walak u \iinchiil. Ma' chenjitn mill, ma' cheti ka mill mak walak u tuklik, u naatik a tukiil alo'. Tulakal; mejen, nukuch, laj ti b'aalo ti walak u winikunb' ul umen u vaj ch' isajil, u men u yaj na , u tat. Jun tiuil kris tic/a no tan u kaal winikil, siempre walak ilik u kaxik jun tuul mak u t-'eetite'; mak u t;'aj it janal, u yuk'itlu. Laj, tulaka/o'on, laj muni tak to'on ti b'aalo. Ma' chcn, "K' u ka b'oob'e'?" "K'u ka b'aalo tun b'etik?" ma' ko'onaki. Bel ti a kuxtala. Dion ilik u t:'aj ti b'aalo. "This is exactly the way il always happens, exactly the way it always comes to pass. It's not just one person, it's not just two people who think, who ponder this way. All, large and small, they have all been made into men this way by those who raised them, their clear mothers and fathers. A person who is coming to manhood, he always looks for someone who will feed him, someone to make his food, his drink. All, all of us, we have all passed that way. It's not just 'What's this?' We can't say, 'Why are they doing this?' This is the nature of life." God himself has made it this way.
The girl's family finally succumbs to the logic of this position, which is accompanied by profuse assurances that the prospective groom's parents (and especially his mother) will be kind to the young girl and patiently help her to complete her domestic education. Ma' tiktttb'atan in jok'xik; wajtub'u xavb'ej, waj titb'axookpokche'. T:'i tinjok'sik ka xiik ti kuntal te'i; siempre ichil in t:' i hatch, in t:' i jedeeb' al u b'el. T:aj toj ilik in tojkintikoo', T:aj toj ilik it kiinanb'aanaloo'. Siempre leek ilik a nooch ch'up, b'el it kit t;'i ilaj, b'el it kit t:' i kiinante' tun. "I will not leave her just anywhere, at some crossroads or in some hollow tree. I will humbly take her to live there, just in my home, in my resting place. Of course I will teach them well. Of course they will be well taken care of. My wife herself will watch over her, she will care for her."
Creating Tzik Relationships
59
There are very few older Mopan people of traditional lifestyle who have never married. The End of the Asking At this point preparations begin to be made for a fourth visit of the young man's family to the house of the young girl. This is called in Mopan 'the end of the asking' (job'ol n k'aat). Both the boy's and the girl's family recruit female labor for this event, since each side will feed the other. The girl's parents and godparents receive the boy and his parents and godparents in the front part of their house, while a frenzy of food preparation activity takes place in the kitchen. An exchange of kichpan t'an, which may be of astonishing length (half an hour or more), is made among these principals (the boy himself is not expected to participate) and especially between the parents of the bride and the groom, who are to become komaade and kompaade to one another at this event. The formal speech exchange concludes with an exchange of tzik, which consists of the first occasion of mutual address in the compadre/comadre relation. "Dioos komaade," says one participant solemnly to the other, extending one hand. "Dioos kompaade" is the equally serious reply as the two shake hands. By this time, many members of the boy's family will have arrived at the girl's house. They have brought heaping containers of luxury food, and a box containing the gifts of the boy to the girl.6 Late in the day, the groom's present to the bride is opened, and she may retire to try on some of her new finery, which she displays to the guests. The entire gathering, by this time very large and consisting of many members of both the boy's and the girl's family, is fed from both boy's and girl's dishes. Large quantities of rum circulate, and there is probably music (either traditional marimba or taped reggae). Dancing begins, and the merriment continues until the last man must be carried home. The Tz'aj Tzik Ceremony At some point in the festivities it becomes noticeable that members of both families are arranging themselves to be introduced one by one to the newly engaged couple. In each case, as the guest approaches the couple, the parents and godparents of the bride and groom confer among themselves to decide upon the new relationship that the guest will undertake with the young spouse. When the decision has been reached, it is conveyed to the protagonists, who extend their hands and greet one another for the first time in their new capacity. "Dioos mu ," one hears uttered solemnly and shyly; "Dioos mu'" is the self-conscious reply. The timbre of both voices bears witness to the solemnity of this moment. This is known as the occasion of the tz'ajtzik, the 'giving of respect'. More practiced individuals accompany the simple introduction with a short speech. One friend explained to me that this speech explains to each person that you "would like to respect them as they should be respected." She translated a short segment for me: "Up to now I have called you Auntie, but now I will respect you more than Auntie. I will call you Mother-in-law."
60
RELATIVELY SPEAKING
Note that in this example, as in many other cases, the individuals who acquire a new tzik-greei'mg status with respect to one another had actually previously maintained a different greeting relationship, on the basis of other ties (in this case, a relatively distant consanguinity). Because of the fact that fz/£-greeting relationships do not necessarily follow a transitive logic (Danziger 1996a), this observation is not incompatible with the strict requirement that the new husband and wife should themselves not greet one another with a relationship term. Taped performances of this kind of introduction that I have obtained invariably also refer to it as the conferral of a new degree of 'respect' (l:ik): PCX ma' tach ti-t:ik-ik-0 ti b'aji/i. well never 1 A_PL-respect-TR_INC-3B lA_PL-self Well, we never used to respect (t:ik) one another. Pes uleeb'e, ud-o'on icli a jun-p'eel u xutn. well now arrivc_herc-1 B_PL in DET one-CL DET change But now we have come to a turning point. Pes ma yan-0 b'ikij. well NEC exist-3B how There's no other way. Pes t:aj wal tun ti -t:'i- tiik-ik -0 ti-h'ajil. wellOBLIG DUB EMPH lA_PL-HUM-respect-INC-3B !A_PL-sclf We just have to humbly respect (t:ik) one another. After the engagement ceremony has taken place, it is extremely difficult, not to say excruciatingly embarrassing, to call off the match. The reason cited is primarily the expense that the boy's family has already undertaken in providing a feast and making over their gifts to the bride. For example, I was told the autobiographical story of one intrepid young girl who, screwing up her courage to refuse an engagement that her parents had already accepted, put off the announcement of her refusal until the very moment when the knife was being sharpened to kill the pig for the engagement party. At that crucial point, however, she felt she would be committed if she waited even a moment longer, and so made her move. In another case, a match is said to have been broken off after the engagement ceremony, at the forced instigation of the groom's family (he had run off with another girl between the engagement and the wedding date). The original girl kept her gifts, but the tzik introductions that had been made were not honored afterward. The groom's mother in this case was considered much to be pitied for her humiliating ordeal. Divorce and remarriage are not permitted among the mostly Catholic Mopan. Marriages, however, often go awry, and individuals may take new partners informally. But in such situations, Mopan honor the relationships that were first constituted through formal introduction at the job'ol u k'aat ceremony. These relationships do not lapse, whatever the status of the particular union that was the occasion for their original creation.7 Conversely, new relations arising from sponsorship of any
Creating Tzik Relationships
61
kind are never undertaken without a similar introduction ceremony being performed, in anticipation of a church marriage (cf. Danziger 1996a). Those who undergo informal marriage (see earlier discussion) do not acquire relationships of this kind. Performative Aspects of Tzik Appellation I was present at one baptism celebration at which the young parents and godparents solemnly decided, after some deliberation, that since they were already related as suku'un/kiik/itz'iin 'same generation consanguines/brothers and sisters', they did not choose to become komaade and kompaade to one another at all. These were young people, and their motivation was to avoid the strict traditional formality of interaction that ideally accompanies Mopan relationships of compadrazgo. Their decision was accepted, albeit in a spirit of disappointment and dismay, by all present, including the elders. No introduction ceremony was performed. Neither was kichpan fan exchanged, despite the fact that some of the older people had been rehearsing their performance in solitary anticipation of the alliance. The two families are today not considered to be in compadrazgo relationship, although the godparent relation itself, cemented in a church ceremony, is recognized. The dilemma posed by the potential doubling up of tzik relationships is also well recognized by Mopan. A woman who has always been addressed by her younger first cousin as kiik 'senior same-generation consanguine/elder sister', for example, finds that he is to marry a girl to whom she is already the godmother. She is now entitled to become all' 'different-generation affine/mother-in-law' to the boy. The individuals concerned in situations like this invariably choose one or the other of their possible tzik statuses relative to the new spouse and thenceforth are considered to occupy only that one. My friends and acquaintances joined in spontaneous debate on this sort of topic more than once during my stay, reviewing the benefits of remaining in the relationship that one was 'used to' (suk-tal) as against the increased respect that would come with the adoption of the new status. Either option was considered a defensible one, but if the new status was not chosen and no introduction made, the individuals concerned were never considered, by themselves or anyone else, actually to be related in the manner not chosen. Rejection of a tzik status to which one is entitled can take place for less friendly reasons. In general, individuals who are related through biology or ritual sponsorship, but who do not present themselves at the appropriate ceremony of tz'aj tzik introduction, are not considered to have entered into new tzik relations with one another as a result of the marriage. This kind of absenteeism therefore constitutes a deliberate and highly visible repudiation of the alliance that is being celebrated.8 For these and other reasons, it sometimes happens that the closest biological relatives are not present at engagement and other tz'aj tzik events. When this occurs, there is no scientific or essentialist concern that the missing parties should be sought out and identified. The concern, rather, is that a gathering may supply too few actually present candidates for tzik relationship to the new couple. Under these circumstances, doubt-
62
RELATIVELY SPEAKING
fully distant biological relations may be pressed into service. These are unhesitatingly categorized as legitimate tzik relations after the ceremony. In the case of relationships acquired through sponsorship and marriage, then, this first public greeting constitutes a case of linguistic usage like those described by Austin (1979 [ 1961 ]) as "performative," in which the act of speech itself makes true the fact of reality to which it refers. Many Mopan consider this the focal moment of all tzik exchange.9 Negotiating Tzik Class Assignment Once willing candidates for new tzik greeting status have been identified, a new problem arises. Exactly which of the possible terminological statuses will the new relationship occupy? Occasions of tz'aj tzik are often the scene of discussion and argument as issues of essential relationship between the two parties are negotiated. The discussions are sometimes quite heated, since individuals may have very different opinions on the subject. What often is at issue is the question of what I have been calling "generation," since affinal relations (like those among consanguines) are terminologically stratified in terms of the age and experience of the two parties with respect to one another, rather than in terms of strict genealogical calculation. Thus the Mopan term previously glossed as 'mother-in-law' (